System for monitoring the operation and maintenance of industrial equipment

Abstract

The invention relates to a system (1) for monitoring the operation and maintenance of equipment (2) in a facility (3), the equipment (2) being operated until a scheduled maintenance (6) is performed and then at least one subsequent scheduled maintenance (7) resulting in the creation of a production and maintenance log (9), a usage log (11) and a log (120) of conditions (13). By characterizing the tasks (90) and conditions (110) that influence the conditions (13), correlations (14) between causes and consequences of the aging of the equipment (2) are determined. For the other instruments, the data corresponding to that correlation (14) is recovered and extracted to train a virtual model (16). The day before the maintenance (6) is to be performed, based on the logs (9, 11), the tasks (90) of the maintenance (6) to be performed, and the scheduled conditions (110) of the scenario (8), the model (16) generates a scheduled state (130) to be compared with the minimum operating state (17) of the equipment (2).

Description

[Technical field]

[0001] The present invention is in the field of industrial equipment monitoring for the purpose of managing the life cycle of industrial equipment. - Optimizing its production and adapting it to a given usage profile, - Optimize and adapt its conservation to the intended usage profile, - determining its maintenance thus optimized over the long term of its service life, - Identify the usage amount that matches the absence of failures during operation between maintenance intervals (zero failures), - Identify the amount of usage that will meet the zero failures requirement over a long period of its useful life; - Direct and optimize its service life extension for the intended usage profile.

[0002] The invention finds preferred, but not exclusive, application in the monitoring of medium to large industrial equipment and installations, in the sense that the equipment in question is subject to reliability requirements, all the more so when reliability requirements are combined with safety requirements. - equipment, the non-availability of which disrupts the service or production of the entire facility to which it belongs; - Equipment, the non-availability of which would cause or aggravate an accident or contingency situation at the facility.

[0003] The invention therefore finds non-limiting application in the transport sector (road, aviation, marine, rail), but also in the process industry sector (utilities for "community services" such as the production of electricity, water, fuel, etc.) or in the defense sector.

[0004] Thus, in the context of the present invention, the general term "equipment" encompasses components of an installation, ie components that are functional and essential for the correct operation and / or safety of this installation.

[0005] Thus, the equipment of interest may consist of components of a mobile facility, such as a machine or vehicle (e.g., a submarine, ship, aircraft or spacecraft), or components of a fixed facility, such as a structure or infrastructure (e.g., a power plant, an oil platform or a refinery).

[0006] The invention therefore meets the high expectations of industrial actors involved in the value chain of this type of equipment, in particular designers, manufacturers, maintenance engineers and operators, since it allows for excellent reliability assurance in terms of production, maintenance and operation.

[0007] The reliability of the equipment in question is important because its failure could result in the halting of production at the facility, and even more so when its reliability is a safety requirement of the facility to which it belongs.

[0008] In this regard, the present invention makes the following choices: The reliability of an instrument depends on the quality of its manufacture, its past maintenance, and its amount of past use, as well as the amount of future use for which the instrument is intended.

[0009] Thus, the issue of equipment reliability illustrates all the benefits of controlling how the design and manufacture, maintenance and usage of equipment affect its reliability.

[0010] It is therefore essential to answer the following questions: - What tasks (gestures, parts, and settings) for manufacturing a device will optimize the device's behavior during operation, taking into account its intended usage? - Do the tasks to be performed for the maintenance of the equipment allow the optimal ratio between the maintenance costs and the probability that the equipment will have zero failures until the next maintenance (taking into account the future usage for which the equipment is intended and taking into account its own production, maintenance and usage logs)? - What is the limit of equipment usage observed between maintenance intervals, i.e., what is the maximum usage that is compatible with zero failures between maintenance intervals?

[0011] It is also useful to constantly determine the optimal tasks for each of the maintenance operations in the remaining useful life of the equipment, as well as its remaining useful life (taking into account future usage and its own logs). In fact, this allows: - Customizing maintenance schedules for each device in a series; - Refreshing and forecasting spare parts provisioning (nature of parts and replacement dates); - Refresh and forecast industrial maintenance scheduling; - Refresh and forecast the total cost of future maintenance over the equipment's useful life.

[0012] It is also always useful to determine: - a limit on equipment usage that is consistent with zero failures observed during each of the future operating periods in the equipment's life; - The remaining useful life of the equipment (taking into account its future usage and its own logs).

[0013] In practice, this allows for realistic planning of the remaining useful life of the equipment and therefore the amount of usage that can be achieved over the long term.

[0014] Furthermore, the ability to refresh and forecast the equipment's usage limits for the remaining useful life, as well as the total cost of future optimal maintenance work, makes it possible to constantly determine the equipment's residual value in terms of possible resale.

[0015] In summary, the issue of being able to control the effects of equipment manufacture, maintenance, and usage on equipment reliability is important for at least the following reasons. - Unlocking access to manufacturing and maintenance optimization; - Facilitate dialogue with industrial conservation workers; - unlocking access to control the reliability of the equipment (by removing uncertainty about its reliability during operation); -Releasing access to control and optimization of the management of the asset that the device represents. [Background technology]

[0016] Currently, the existing methods that can have the greatest impact on equipment reliability are: - Predictive maintenance, - physical model, - Certification programs on the test bench, - Feedback (commonly abbreviated as "FB").

[0017] However, these methods do not provide the simulation necessary to control the impact of the equipment's manufacture, maintenance, and usage on its reliability, and in this sense they only partially meet the manufacturers' expectations.

[0018] With respect to predictive maintenance, the technique observes and analyzes the operation of equipment in the hope of detecting deterioration. Once deterioration is detected, a first trend of the deterioration's progression is constructed to display a first prediction of the future failure date. This trend of progression and this failure date are then iteratively fine-tuned with subsequent observations.

[0019] In so doing, predictive maintenance is subject to several methodological biases, namely: - seeks to extrapolate the future behaviour of equipment from its observed past behaviour and therefore the future consequences of ageing, independent of the past and future causes of ageing; - Assuming future usage will be the same as past usage. This is almost always false, - Use only operational data, regardless of functional evaluation.

[0020] Therefore, in predictive maintenance, - The first forecast is vague and slow, requiring observation times of the order of 15 days of operation after the maintenance work is performed; - It takes time to show fair predictions, - Myopic, i.e. its forecast horizon is limited to an intermediate time frame of the order of 6 months at most.

[0021] Another consequence of its methodological bias is that predictive maintenance is not proactive enough. In fact, it only reacts to the appearance of deterioration during operation, rather than predicting the appearance of deterioration. Moreover, it requires the operation of the equipment to formulate its predictions, which is an unacceptable operational drawback. More generally, predictive maintenance is performed without the use of any simulation, in particular the simulation of the behavior of the equipment according to its manufacture or maintenance and the planned usage of this equipment.

[0022] Therefore, predictive maintenance is limited to only providing predictions of failure dates with insufficient accuracy and proactive measures, and does not have the ability to provide recommendations regarding production, maintenance or usage to optimize equipment behavior, particularly with the goal of hastening failure dates or extending useful life.

[0023] With the increase in computational means and the increasing value of data, various predictive maintenance solutions have been developed, but they only utilize operational data, are affected by the same methodological biases and, as a result, suffer from the same limitations, and there is little difference between them.

[0024] Regarding physical models, from a technical and especially economic point of view, a complete and customized modeling of a given device by a physical model is not realistic. These models are very expensive and cover an overly restricted field of physical phenomena. Moreover, these physical models are vague.

[0025] In fact, a physical model can only characterize the behavior of the basic parts of the components of the equipment exposed to physical phenomena. Moreover, this modeling is an approximation of reality and contains elements of inaccuracy. In the context of fundamental research, the development of a physical model also requires considerable resources in terms of time and cost. In addition, it is also appropriate to generate a large number of physical models, in the order of hundreds, to characterize the aging of each of the basic parts of the equipment before achieving a global modeling of the equipment.

[0026] Therefore, modeling the behavior of a series of standard equipment by physical models requires considerable resources in terms of cost and time, and in addition, this global modeling based on these physical models is difficult to customize to the case of each equipment, especially its own complete log.

[0027] Regarding the bench tests carried out during the design and especially during the qualification of a given series of benchmark equipment, these tests theoretically make it possible to approximate the behavior of the series of benchmark equipment over its entire useful life with respect to one or more constraints simulating usage and ageing. These tests only give vague indications, since they give trends and magnitudes.

[0028] Irrespective of the specificities of a given series of equipment carried out individually (i.e. without taking into account the manufacturing, maintenance and usage logs of a given equipment, which these tests cannot actually predict), these tests are in fact unable to characterize the future behavior of this equipment as a function of the future amount of use that is to be made, and in particular are unable to determine the parameters of the maintenance to be carried out that will make it possible to optimize the behavior of this equipment during operation.

[0029] In practice, bench testing consists of characterizing the behavior (e.g., steady-state operating point or response to transient conditions) of benchmark equipment of a series of interest when subjected to one or more given usage and aging profiles, and sampled equipment representing the series is then subjected to accelerated aging that is assumed to be representative of real aging. In this case, the characterization of the behavior remains ambiguous, since the aging during the test is accelerated and the usage considered in the course of the test is separate from the actual usage that occurs in the course of the life of each of the equipment in the series.

[0030] The behavior of the benchmark equipment in the series, and in particular its service life, is therefore only an approximation.

[0031] Feedback, generally abbreviated as "FB", complements the trends and orders of magnitude approximated by bench tests. FB also cannot take into account the specifics of a given series of equipment, especially its complete logs, and therefore cannot characterize the future behavior of the target equipment in response to a given future usage, and in particular cannot determine maintenance parameters that optimize its operating behavior. Moreover, data from FB is insufficient due to unsystematic and poorly regular operation.

[0032] In fact, FB makes a comparison between average behavior, such as service life, and average usage. With this approach, precise data cannot be obtained.

[0033] Furthermore, FBs are not used systematically. When FBs are used, it is done very irregularly and the FB refresh frequency is very low, with a periodicity of about one to two years.

[0034] The object of the present invention is to overcome the drawbacks of the prior art by proposing a digital system for monitoring the operation and maintenance of at least one industrial device in a facility, capable of generating and using behavioral models specific to a series of devices.

[0035] The generated behavioral model has relevance, - Correlate the causes and effects of equipment aging, - Specific to the physical phenomenon that each instrument in the series hosts, - Can be customized for each device in the series, - have the accuracy of empirical models that are trained on observed data ("data-driven" models).

[0036] The generated behavioral model is powerful and allows one to simulate the behavior of a series of equipment for a given maintenance decision and a given amount of equipment planned usage, especially taking into account the equipment's manufacturing, maintenance, and usage logs. The model is even more powerful because it allows this type of simulation over an extended period of the equipment's useful life.

[0037] Thus, for a given series of devices of interest, taking into account the intended planned usage of the device as well as its production, maintenance and usage logs, the invention makes it possible to obtain an overall value proposition, in particular: For a given maintenance step of an equipment, the system allows to determine an optimal maintenance decision that allows an optimal balance between maintenance constraints and the probability that the equipment will have zero failures until the next maintenance and for the intended planned usage. This optimal decision is further customized to the equipment's manufacturing, maintenance and usage logs. For a period of operation of the equipment in the course of its useful life, the system makes it possible to determine the limit usage, i.e. the maximum possible usage that meets zero failures until the next maintenance. The system makes it possible to refresh the usage limits of the equipment during operation of the facility, taking into account the actual usage since the last maintenance of the facility and taking into account the intended scheduled usage until the next maintenance. The system likewise makes it possible to constantly determine the schedule of the optimal future maintenance of the equipment during the remaining of its useful life, as well as the limit usage of the equipment observed for each operating period during the remaining of its useful life. - The system always determines the end date of the equipment's useful life. - Finally, the system makes it possible to determine the manufacturing tasks (gestures, parts, settings) that optimize the lifespan of the series of benchmark equipment for a given planned usage scenario, in particular a scenario that depends on the usage segments that intersect with the geographical segments of usage.

[0038] To do this, the invention considers the scale of the series of equipment. For a given series, the invention first determines the correlation between the causes and consequences of the aging of the equipment of the series in question. This correlation links the manufacturing and maintenance logs as well as the usage logs of the equipment of the given series with the condition of this equipment created by these logs. The determined correlation is specific to the physical phenomenon hosted by each equipment of the series. Secondly, the invention models this correlation in the form of a virtual model trained using the manufacturing, maintenance, usage and condition data of the equipment of the series in question. In a third step, the invention uses this model as a customized simulator for each equipment of the series, which in fact makes it possible to simulate the condition of the equipment at a given time for a maintenance decision and for the usage that continues up to this time, taking into account the manufacturing and maintenance logs as well as the usage logs that are specific to the equipment and that precede the maintenance of the target. This model allows access to the value proposition of the invention.

[0039] Moreover, by modeling the behavior of standard equipment in this way, the present invention correlates, on a series scale, the observed behavior of equipment during the manufacture, maintenance, and usage of each piece of equipment. The present invention is thus a method that uses feedback that derives its superior properties from the functional relevance of the modeled correlation. The present invention also makes it possible to use systematic feedback in a continuous or quasi-continuous manner by resuming the training of the model from refreshed manufacture, maintenance, usage, and condition data of the equipment in the series in question, according to a periodicity determined to be optimal by the designers, manufacturers, maintenance engineers, and operators.

[0040] To this end, the invention relates to a digital system for monitoring the operation and maintenance of at least one industrial device in a facility, executed by at least one computing terminal.

[0041] Advantageously, the system comprises at least - installing, in the facility, at least one industrial device resulting from a manufacturing process and representative of a series, and then operating said device at least in relation to the period until the maintenance step to be performed; - defining at least one planned maintenance following said maintenance to be performed after at least one scenario having planned conditions of use of said equipment over a planned operating period.

[0042] Such a system includes a system in which the manufacture, installation, operation, and maintenance of the equipment is performed in accordance with at least: - a production and maintenance log, a task for manufacturing the at least one device to said installation; a production and maintenance log, optionally including at least one proactive maintenance task for the equipment; - a usage log of the device for the period between installation and the maintenance being performed, the usage log including the conditions of use of the device during the period; and - creating a status log for the equipment, the status log including a material indicator for the equipment; - by a technical analysis of the equipment, at least one correlation between at least one of the tasks and / or at least one of the conditions of use and at least one of the material indicators of the condition is determined, the correlation establishing at least one link between a cause of aging of the equipment and a consequence of aging; In correlation, - the tasks are characterized by those identified as critical and / or the conditions of use are characterized by those to which the equipment is sensitive and to which it is exposed during operation or at shutdown, and the tasks and conditions in question affect the state of the equipment; - the material condition of the equipment is characterized by indicators identified as representative of this condition of the equipment; Next, in correlation, - measured physical quantities or functions of measured physical quantities characterizing the conditions of use to which the equipment is sensitive and to which it is exposed during operation or at rest; - a measured physical quantity or a function of a measured physical quantity characterizing the material state of the equipment at a given time is determined; Next, - recovering and extracting, for other devices of the series, the data associated with these tasks and with the physical quantities or functions of physical quantities related to these conditions of use and to these material indicators, identified in the correlation, in order to obtain a data set; - training at least one virtual model based on the dataset.

[0043] The system further comprises: During the maintenance of the equipment in question, at least one value of a task from a production and maintenance log and at least one value of a usage condition from a usage log; - at least one value of the intended conditions of use of the task of the conservation to be performed and the scenario is submitted to the model, The model generates a proposed state of the equipment after the maintenance is performed, and the proposed state is compared to a minimum state identified as necessary for operation of the equipment. [Brief description of the drawings]

[0044] Other features and advantages of the present invention will become apparent from the following detailed description of non-limiting embodiments of the invention, which proceeds with reference to the accompanying drawings. [Figure 1] FIG. 1 shows a schematic diagram of the architecture of a system implemented in monitoring equipment, in particular the submission of the manufacturing and maintenance logs and usage logs of the equipment on the day before the maintenance is to be performed, as well as the planned usage conditions of the maintenance tasks and scenarios to be performed, to a trained virtual model, which generates a planned state of the equipment on the day of the next scheduled maintenance. [Diagram 2] A diagram of the architecture details is shown in schematic form, in particular showing the manufacturing and maintenance logs including the manufacturing decisions of the equipment up until the installation of the equipment, and the proactive maintenance decisions of the equipment, as well as the to-be-performed maintenance decisions and the planned maintenance decisions, with particular emphasis on the tasks of each of the decisions and their timing. [Diagram 3] A schematic diagram of the architecture details is shown, in particular showing the equipment usage log between installation and the maintenance to be performed, as well as planned usage scenarios, highlighting that the usage log and the scenarios include the equipment usage conditions associated with the respective time periods. [Figure 4] 1 shows a schematic diagram of the architecture details, in particular showing a log of successive states of the equipment, as well as successive planned states of the equipment, with particular emphasis on the material indicators of the equipment states associated with their respective time periods. [Diagram 5] 1 shows a schematic diagram of the architectural details of the surveillance system, and in particular the process of recovering and extracting data from one of the series of devices as identified in the trained correlation. [Figure 6] FIG. 13 shows a schematic diagram of another detailed view of the architecture, specifically illustrating the process of recovering and extracting data for multiple instruments in a series, in order to obtain a dataset and train a virtual model. [Figure 7] A schematic diagram of the architecture of the monitoring system is shown, in particular the virtual model generating the planned state of the equipment at a given point in time and its two material indicators, together with a comparison against the minimum state identified as necessary for the operation of the equipment. [Figure 8]1 shows a schematic example of a calculation of the limit of the planned condition of a usage scenario after maintenance has been performed, such that at the next scheduled maintenance, the planned state of the equipment is equal to the minimum state. [Figure 9] Schematically shows another diagram of the example of another calculation of another limit for another planned condition of the usage scenario after maintenance is performed, such that at the next scheduled maintenance, the planned state of the equipment is equal to the minimum state. [Figure 10] In particular, the example diagram is shown diagrammatically with a scale of values ​​of a particular planned use condition in the form of a polygon, showing the target values ​​and maximum limits of the planned condition associated with the scenario. [Figure 11] 1 shows a schematic diagram of such an example having a set of curves on a chart, particularly highlighting the maximum limit of a third of the predetermined conditions from the variation of two of the predetermined conditions. [Figure 12] 1 shows a schematic diagram of a long-term architecture similar to that of FIG. 1, and in particular a first step of implementing the monitoring system recursively to predict planned states from performed maintenance actions to planned maintenance actions. [Figure 13] In particular, it shows a schematic diagram similar to FIG. 12 over a longer period of time showing a second step of implementing the monitoring system recursively to predict a planned state from the scheduled maintenance operation to the next scheduled maintenance operation. [Figure 14] 1 shows a simplified diagram of an example of a long-term forecast, in particular showing a curve representing the progression of the equipment condition as a function of time, in the course of two successive planned time periods; [Figure 15] 1 shows a simplified diagram of an example of a long-term forecast, in particular showing a curve representing the progression of the state of an equipment as a function of time, over the course of several successive planned time periods; [Figure 16] 1 shows a simplified diagram of the architecture of a monitoring system, in particular showing the virtual model for determining the point in time of failure date of the equipment for insufficient maintenance decision. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0045] The invention relates to a system 1 for monitoring the operation and maintenance of at least one device 2 in an industrial installation 3 .

[0046] Such a surveillance system 1 (hereinafter "system 1") is foreseen as digital, in other words it is at least one software program intended to be executed by at least one computing terminal.

[0047] Typically, such a computing terminal may be of any type, in particular a computer server or a computer. Moreover, the computing terminal allows, via suitable storage means, to record, read, modify and generate data in digital form. It is also assumed that the computing terminal is accessible locally or remotely via a suitable communication network.

[0048] In addition, such a system 1 comprises successive steps which are described below in a non-exhaustive manner: Thus, according to the invention, the system 1 can be connected to a method.

[0049] Some steps are implemented in a real manner, in particular via a virtual interface, in particular by a person interacting with the computing terminal in question, or in particular when an operator interacts with the device 2 in question.

[0050] Other steps are performed virtually, in particular when data processing is performed by a computing terminal, so that the system 1 comprises digital elements as well as virtual technical means which are advantageously formed and implemented within the scope of the present invention.

[0051] The system 1 may provide for the monitoring of one of the devices 2 in one facility 3 , the monitoring of multiple devices 2 in the same facility 3 , or even the monitoring of multiple devices 2 in multiple facilities 3 .

[0052] As previously mentioned, the equipment 2 is characterized by being essential to the operation of the facility 3. As such, the unavailability of the equipment 2 may interrupt production at the facility 3 and may cause or exacerbate an accident situation at the facility 3.

[0053] The device 2 may be of any type, for example a heat engine, an electric motor, an alternator, a pump, a solenoid valve.

[0054] The equipment 2 therefore forms an integral part of the installation 3 and is necessary for its correct operation.

[0055] Such a facility 3 may be mobile, such as a machine or vehicle (e.g. a submersible, ship, aircraft or spacecraft), or may be fixed, such as a structure or infrastructure (e.g. a power plant, oil platform or refinery).

[0056] The system 1 therefore comprises, as an initial condition, the installation in an industrial facility 3 of at least one piece of equipment 2 resulting from a manufacturing process 4 and representing a series, i.e. the equipment 2 has been integrated into the facility 3 in the past.

[0057] Furthermore, the manufacture 4 of the device 2 includes the assembly of several components to obtain the device 2, followed by the installation of the device 2. Prior to this manufacturing process 4, a design 40 of the device 2 is performed.

[0058] It should be noted that equipment 2 represents a series including all equipment 2 identically manufactured from a single design plan, where one piece of equipment 2 in the series may no longer be operational, another piece is currently operational, and another piece already manufactured may remain installed and operational.

[0059] After being installed in the facility 3, the device 2 operates for at least a period 5 up to a maintenance step 6 to be performed. The period 5 therefore corresponds to a period of operation of the device between installation and a maintenance step 6 to be performed, or otherwise corresponds to an alternation of periods of operation and maintenance steps since installation.

[0060] It should be noted that the maintenance steps 6 to be performed are concrete actions that require the intervention of at least one operator on the geographical site of the facility 3 and are performed directly on the equipment 2 in question.

[0061] In this regard, as mentioned above, the monitoring system 1 is assumed to manage the future life cycle of the device 2, i.e. - defining the nature of the future maintenance tasks of equipment 2 in order to optimize them and adapt them to the intended usage profile of said equipment 2, and doing so over a long period of time, i.e. during the useful life of equipment 2; - for each planned operating period 70 of equipment 2, identify a usage limit that meets zero failures during operation until the next maintenance, and do so over the long term and useful life of equipment 2; - Zero failure, device 2 usage limit refreshed in real time during operation; - Prescribing the service life extension period of equipment 2 by defining the nature of the final maintenance tasks of equipment 2 optimized and adapted to the intended usage profile of equipment 2.

[0062] At the scale of a series of equipment, the monitoring system 1 also envisages optimizing the production of standard equipment representing the series in order to determine manufacturing parameters 4 to adapt to a given usage profile, i.e. optimizing the life cycle of the equipment 2.

[0063] Thus, at least one planned maintenance 7 following the maintenance 6 to be performed is defined.

[0064] The future period following between the performed maintenance 6 and the scheduled maintenance 7 corresponds to at least one scheduled operating period 70, during which the equipment 2 operates in at least one scenario 8 characterized by a scheduled use condition 110, referred to as "scheduled condition 110". The future period may also correspond to alternation of scheduled periods of work and maintenance steps.

[0065] The future period may extend to the end of the equipment's useful life.

[0066] FIG. 1 shows, inter alia, the design 40 and manufacturing process 4 of a piece of equipment 2, the time period 5 (between the installation of the equipment 2 in a facility 3 and the maintenance to be performed 6, including any pre-emptive maintenance 10), and the planned time period 70 until the planned maintenance 7.

[0067] In this case, the steps of the manufacturing process 4 and installation of the equipment 2, as well as the operation and maintenance of the equipment 2 during the course of time 5, generate several types of data that the monitoring system 1 takes into account.

[0068] The manufacture, installation, operation, and maintenance of the device 2 produces at least one manufacturing and maintenance log 9 (or "log 9"), the log 9 including: - a task 90 for manufacturing said device 2 until said installation; - optionally including at least one proactive maintenance 10 task 90 of said device 2.

[0069] In other words, the manufacturing and installation steps include at least manufacturing and installation tasks 90. Additionally, the time period 5 optionally includes at least one proactive maintenance 10 of the equipment 2, including maintenance tasks 90. All of the tasks 90 of the at least one equipment 2 up to the installation, as well as any tasks 90 of the at least one proactive maintenance 10 of the equipment 2, generate a manufacturing and maintenance log 9.

[0070] It should be noted that the manufacturing tasks 90 as well as each task 90 of the maintenance operation 10 are defined, in a non-limiting manner, by gestures performed by an operator, by parts removed during a maintenance step that are installed or replaced during manufacturing, and by adjustments of the equipment 2 in question.

[0071] As can be seen in FIG. 2, the system 1 includes, inter alia: a manufacturing and maintenance log 9, i.e. a log of manufacturing and maintenance decisions carried out with respect to the manufacturing process 4 and installation (of rank 0) and with respect to the proactive maintenance 10 (of ranks 1 to j-1); - A conservation judgment regarding conservation 6 of the implementation target (rank j); - Maintenance judgment regarding planned maintenance 7 (rank j+1).

[0072] Each production or maintenance decision of rank k is characterized by its ratio to various possible tasks 90 (optional performance of the task 90, characterization of the task 90 when executed, and timers 91 characterizing the period after which the task 90 is performed on the equipment 2).

[0073] Additionally, this step also creates a usage log 11 (or “log 11”) of the device 2 over the period 5 between the installation and the maintenance 6. The usage log 11 includes conditions 110 (also referred to as “conditions 110”) for using the device 2 during the period 5.

[0074] Indeed, the period 5 between the installation of the equipment 2 and the maintenance 6 to be carried out includes at least one period of operation of the equipment 2 in relation to usage.

[0075] This usage is related to the way the device was used, ie the set of usage conditions 110 for that device 2 during that period.

[0076] For example, in the case of equipment 2 corresponding to an engine of a facility 3, such as a truck, these usage conditions 110 may be the load transported, the distance traveled, the speed, the ambient air temperature, the average gradient of the route used.

[0077] This usage thus generates a usage log 11 for that equipment over the period 5 between installation and that maintenance 6 being performed. As can be seen in FIG. 3, which shows the case of an equipment 2 such as a pump (where one of the conditions 110 is temperature), the system 1 may, inter alia: - a log 11 of the use of device 2 during period 5 (from installation of rank 0 to maintenance of the target of rank j 6); - a scenario 8 for the usage of the equipment 2 forecasted over a planned period 70.

[0078] The usage log 11 can be decomposed into a set of usage sub-logs 111 of the equipment 2. The usage sub-log 111 of rank k represents a portion of the usage log 11 of the equipment 2 in the course of an operating period comprised between two successive proactive maintenance operations 10 of rank k and rank k+1. In particular, FIG. 3 highlights the following: - Rank 0 Usage Sublog 111 (between Rank 0 Installation and Rank 1 Advance Maintenance 10), - Usage sublog 111 of rank j-1 (between the last preceding maintenance work 10 of rank j-1 and the maintenance to be performed 6 of rank j).

[0079] The log 11 includes all of the used sub-logs 111 from rank 0 to rank 1.

[0080] The usage sublog 111 of rank k represents the progression of each of the usage conditions 110 between the two pre-emptive maintenance operations 10 of rank k and rank k+1, such as the progression of temperature over time (as can be seen in FIG. 3).

[0081] In the usage sublog 111 of the equipment 2 of rank k, it is possible to distinguish the progress of each of the usage conditions 110 between the advance maintenance 10 of the rank k and the time (t) for the partial usage sublog 1110 of the equipment 2 of rank (k, t), i.e., for the time (t) included between the maintenance of rank k and the maintenance of rank k+1. Scenario 8 represents the planned progress of each of the planned usage conditions 110 in the context of the planned usage of the equipment 2 between the maintenance to be performed 6 (rank j) and the planned maintenance 7 (rank j+1), as seen in FIG. 3. In the scenario 8, it is possible to distinguish partial scenarios 80. The partial scenario 80 of rank (j, t) represents the planned progress of each of the planned conditions 110 of the equipment 2 between the maintenance to be performed 6 of rank j and the time (t) for the time (t) included between the maintenance to be performed 6 of rank j and the planned maintenance 7 of rank j+1.

[0082] Additionally, this step also creates a status log 12 (or “log 12 ”) for the equipment 2 that includes the material indicator 120 (or “indicator 120 ”) for the equipment 2 .

[0083] Indeed, the amount of usage of the equipment 2 causes aging degradation that affects the condition 13 of the equipment 2. The condition 13 of the equipment 2 at a given time, characterized by material indicators 120, reflects the physical integrity of the equipment 2 on which the operational capability of the equipment 2 depends. All conditions 13 of the equipment 2 (or all material indicators 120) that actually occurred during the time period 5 constitute the condition log 12 of the equipment 2.

[0084] As can be seen in FIG. 4, which illustrates the case of equipment 2 such as a pump (one of the material indicators 120 is the flow rate (Q)), the system 1 includes, among other things, a status log 12 of the equipment 2 during a period 5. The status log 12 can be decomposed into a set of status sub-logs 121. The status sub-log 121 of rank k of the equipment 2 represents a portion of the status log 12 in the course of the operating period comprised between two successive proactive maintenance operations 10 of rank k and rank k+1. In particular, FIG. 4 highlights the following: - Rank 0 status sublog 121 (between Rank 0 installation and Rank 1 advance maintenance 10), - State sublog 121 of rank j-1 (between the last preceding maintenance 10 of rank j-1 and the maintenance to be performed 6 of rank j).

[0085] The log 12 includes all the status sublogs 121 from rank 0 to rank j-1. As can be seen in FIG. 4, the status sublog 121 of rank k of the equipment 2 represents the progress of the successive statuses 13 of the equipment 2 between the two preemptive maintenance operations 10 of rank k and rank k+1. The status sublog 121 of rank k of the equipment 2 represents the progress of each of the material indicators 120 (e.g., the progress of the flow rate over time) between the two preemptive maintenance operations 10 of rank k and rank k+1. As can be seen in FIG. 4, in the status sublog 121 of rank k of the equipment 2, it is possible to distinguish partial status sublogs 1210. The partial status sublog 1210 of rank (k, t) represents the progress of each of the material indicators 120 between the preemptive maintenance 10 of rank k and the time (t) (for the time (t) between the maintenance of rank k and the maintenance of rank k+1).

[0086] Thus, logs 9, 11, 12 extend in time during period 5. They contain production and maintenance tasks 90, use conditions 110, and material indicators 120, respectively.

[0087] The aforementioned elements of the logs 9, 11, 12 represent the specification of calculated fields in which the measurements coming from the device 2 in question are recorded continuously over time.

[0088] Advantageously, in a first step, at least one correlation 14 is determined between at least one of said manufacturing and maintenance tasks 90 and / or at least one of said conditions of use 110 and at least one of said material indicators 120 of said state 13. Said correlation 14 establishes at least one link between causes of ageing of equipment 2 and consequences of ageing.

[0089] In other words, the present invention chooses to approximate the behavior of the equipment 2 in terms of the causes and consequences of ageing.

[0090] For this reason, the invention chooses to characterize the causes of ageing of the equipment 2 by a manufacturing and maintenance log 9 of the equipment 2 and by a usage log 11 of the equipment 2. The invention further chooses to characterize the consequences of ageing of the equipment 2 by a state 13 of the equipment 2 caused by said logs 9 and 11.

[0091] Therefore, the present invention chooses to correlate the status 13 of the equipment 2 with the manufacturing and maintenance log 9 and the usage log 11 of the equipment 2 .

[0092] Thus, the present invention foresees correlating a given manufacturing and maintenance log 9 (hereinafter referred to as the “first term” of the correlation 14) and a usage log 11 (hereinafter referred to as the “second term” of the correlation 14) of a series of equipment 2 with a state 13 of this equipment 2 caused by these logs 9, 11 (hereinafter referred to as the “third term” of the correlation 14), said three terms of the correlation 14 constituting a triplet 140.

[0093] For example, consider the case of a pump-like device 2 that is sensitive to the temperature of the fluid being delivered. The present invention correlates the pump's usage log (characterized by a fluid temperature log and / or a suction pressure log and / or a pump speed log), as well as the pump's main manufacturing options (such as the type of impeller installed) and the pump's maintenance log (such as a log of impeller replacements during various proactive maintenance operations 10), with the pump's condition (characterized by its flow rate and / or discharge pressure).

[0094] Furthermore, the correlation 14 is intended to take into account only relevant parameters that are likely to affect the ageing and operation of the equipment 2, other parameters being less suitable for selection.

[0095] To this end, a technical analysis of the equipment 2 is performed to identify the critical manufacturing and maintenance tasks 90 and use conditions 110 that affect the status 13 of the equipment 2 .

[0096] To do this, through an engineering analysis of the equipment 2, the manufacturing and maintenance tasks 90 are characterized by tasks that are identified as being critical, as defined below.

[0097] More precisely, the invention chooses to characterize the tasks 90 of the manufacturing and maintenance log 9 of the equipment 2 in at least one of the following ways: - for a manufacturing process 4 of equipment 2, the invention characterizes the gestures set, the parts installed, or even the adjustments adopted during the manufacturing process. The gestures of the manufacturing process are characterized by a protocol option for each manufacturing gesture, if there are several possible protocol options for said gesture for the manufacturing process 4 of the equipment 2 of the series. The parts installed during the manufacturing process are characterized by the option of the part installed, if there are several possible options for said part for the manufacturing 4 of the equipment 2 of the series (e.g. if the manufacturing of a pump of the series foresees two possible types of pump bearings). The adjustments adopted during the manufacturing process are characterized by the value of each adjustment, if there are several possible values ​​of said adjustment for the manufacturing 4 of the equipment 2 of the series (e.g. the tightening torque of a pump gland). - for the maintenance steps of the equipment 2, the invention characterizes the gestures set, the parts installed, or even the adjustments adopted during each maintenance step. Maintenance gestures include replacing parts of other maintenance tasks 90 (e.g. re-tightening a connection of an electrical terminal block). Replacement of parts during the maintenance of the object is characterized by the option of the part if there are several possible options for said part for the maintenance of the equipment 2 of the series (e.g. if a pump in the series foresees two possible types of pump bearings). Adjustments adopted during the maintenance of the object are characterized by the value of each adjustment if there are several possible values ​​of said adjustment for the maintenance of the equipment 2 of the series (e.g. tightening torque of a pump gland).

[0098] Additionally, according to one embodiment, in the correlation 14, the production and maintenance log 9 is reduced to a log of critical tasks 90 in the form of at least one list of consecutive values.

[0099] In other words, through a technical analysis of the equipment 2, the present invention chooses to characterize the manufacturing and maintenance log 9 of the equipment 2 by taking into account only the critical tasks 90, i.e., gestures, parts and adjustments identified as determining the behavior of the equipment 2 during operation (i.e., as affecting the aging of the equipment 2 and therefore the state 13).

[0100] In addition, the manufacturing and maintenance log 9 of the equipment 2 (up to and including the last proactive maintenance action 10) is characterized by enumerating, for each critical task 90, the values ​​adopted successively during the manufacturing process 4 and during the various successive proactive maintenance actions 10 in the life of the equipment 2, up to the last proactive maintenance action 10 (the one preceding the maintenance to be performed 6). The manufacturing and maintenance log 9 is therefore characterized by a list of values. Consider, for example, the case of an equipment 2 such as a pump, which comprises bearings and the placement or replacement of pump bearings and the nature of which correspond to the critical manufacturing and maintenance tasks 90 of the pump. In this example, the last proactive maintenance action 10 (the one preceding the maintenance to be performed 6) is considered to be the seventh (rank k=7). The invention characterizes this log of critical manufacturing and maintenance tasks 90 by a list [A,0,0,B,0,0,B,0], indicating the placement of the bearing of type A in the manufacturing operation 4, the replacement of the bearing with a new bearing of type B in the third maintenance operation, the replacement of the bearing with a new bearing of type B in the sixth maintenance operation, as well as the fact that no maintenance work has been performed on the bearing in the other predecessor maintenance operations 10. The invention therefore chooses to characterize the manufacturing and maintenance log 9 of the equipment 2 by a matrix made up of lists, each list associated with one critical task 90, enumerating successive values ​​that characterize this critical task 90 according to the manufacturing process 4, then enumerating the various maintenance operations of the equipment 2 up to and including the last predecessor maintenance operation 10. For example, consider the aforementioned case of equipment 2 such as a pump, where the manufacturing and maintenance log 9 can be reduced to logs for two critical tasks 90 "Pump bearing placement or replacement" and "Pump impeller placement or replacement", with lists [C,0,A,0,0,B,0,C] and [A,0,0,B,0,0,B,0] as the respective log lists after the seventh maintenance operation. Then, the manufacturing and maintenance log 9 of the target pump up to the seventh maintenance operation corresponds to the matrix [[C,0,A,0,0,B,0,C];[A,0,0,B,0,0,B,0]].

[0101] In addition, the use conditions 110 are characterized by conditions to which the equipment 2 is sensitive and to which it is exposed during operation or at rest.

[0102] In fact, by technical analysis of the equipment 2, the invention reduces the conditions 110 to ambient conditions and / or operational conditions (CA / CF) to which the equipment 2 is sensitive and to which it is exposed during operation or at rest.

[0103] Ambient conditions (CA) are understood to be conditions of the environment external to the equipment 2 to which the equipment 2 is sensitive and to which it is exposed during operation or at rest (ambient air temperature, hygroscopicity, irradiance, etc.).

[0104] It should be understood that the operating conditions (CF) are as follows: - the conditions inside the equipment 2 to which it is sensitive and to which it is exposed during operation or at rest (such as the temperature of the fluid being conveyed, the level of vibration in the case of a pump, etc.); and / or - parameters describing the power deployed by the equipment 2 during operation and influencing its operating point (load to be transported, speed in the case of a truck engine, etc.), and / or - other parameters necessary to characterize the total amount of work that equipment 2 has undergone under the aforementioned ambient and operating conditions (CA / CF) (such as the total distance traveled in the case of a truck, or generally the usage hours of equipment 2).

[0105] Consider, for example, the case of equipment 2 such as a pump, the impeller of which is made of a thermoplastic material and is therefore sensitive to and exposed to the temperature of the conveyed fluid, in which case the temperature of the conveyed fluid is a much more important operating condition to take into account than in the case of pumps whose impellers are made of metal.

[0106] Within the meaning of the present invention, the scenarios 8 and the intended use conditions 110 are characterized by the same ambient and operating conditions (CA / CF) to which the equipment 2 is sensitive and exposed during operation or at rest.

[0107] Additionally, the material condition 13 of the equipment 2 is characterized by a material indicator 120 that is identified as representing this condition 13 of the equipment 2 .

[0108] To do this, through technical analysis of the equipment 2, the present invention selects to characterize the current state of the equipment 2 by a necessary and sufficient set of performance, vibration behavior, and other material indicators 120 (e.g., without limitation, material indicators typically measured by non-destructive testing techniques) that are determined to be representative of the condition 13 of the equipment 2.

[0109] In order to characterize the correlation 14, and in particular the log of conditions 110, as specifically as possible for the equipment 2 in question, a measured physical quantity or a function of a measured physical quantity is determined which characterizes the conditions 110 to which the equipment 2 is sensitive and exposed during operation or at rest, or which best characterizes the log 110 of these usage conditions (i.e. the usage log 11).

[0110] These functions of physical quantities include multiple mathematical or algorithmic functions. In other words, each condition 110 has a measured physical quantity associated with it that characterizes this condition. Thus, where relevant, the invention characterizes the log of a condition by the log of this physical quantity.

[0111] Thus, for example, if the equipment 2 is sensitive and exposed to the ambient air temperature or, in the case of a pump, the temperature of the fluid being conveyed, the log of the corresponding temperature is taken into account. Furthermore, the invention also chooses to characterize the log of the condition 110 by the log of a function of a physical quantity representing the condition 110, when this characterization is more appropriate than the log of this physical quantity.

[0112] For example, if the equipment 2 is sensitive and exposed to the ambient air temperature or the temperature of the fluid being transported (as in the case of a pump), it is possible to integrate the log of temperature to the time integral of the temperature over the period between the installation of the equipment 2 in the facility 3 and that time.

[0113] Furthermore, when the present invention chooses to characterize the log of a condition 110 by the log of a function of a physical quantity representative of the condition, these functions can include, without limitation, a calculation of the duration of the measured physical quantity over at least one range of values.

[0114] In fact, the present invention chooses to consider that a collection of abnormal transients influences the aging and thus the behavior of the equipment 2. Therefore, where relevant, the present invention characterizes the log of a condition 110 by counting the number of occurrences of this condition within each of the normal operating range, the near-destructive range, or an intermediate range between the previous two.

[0115] Consider, for example, the case of equipment 2 such as a pump that is sensitive and exposed to the temperature of the fluid it conveys. The log of temperature can then be characterized by the time integral of the temperature over the period from installation of the equipment 2 in the facility 3 to a time point, and components of said integral that are within the normal operating range can be distinguished from components that are in a range close to destruction, from components that are in an intermediate range between two preceding regions. If the invention chooses to characterize the log of a condition 110 by the log of a function of a physical quantity that represents said condition, these functions can also include, without limitation, a calculation that represents the variation of at least one of the measured physical quantities, and / or a count of said at least one variation.

[0116] This is especially the case when the instrument 2 is sensitive and subject to variations in the conditions 110. Without limitation, such a function can correspond to a gradient function or a count of cycles of the condition of interest.

[0117] A first example would be the case of equipment 2, such as a pump, that is sensitive to and subject to temperature of the fluid being conveyed, and in particular to sudden changes in that temperature. To characterize the log of conditions 110 up to a given time, the average slope of this temperature during a temperature transient (i.e., during a transient that causes a sudden change in temperature) can be calculated (e.g., an average change of 20°C / min "degrees Celsius per minute"). This average slope can then be correlated with a count of similar transients (e.g., 2000 sudden temperature transients correlated with an average slope of 20°C / min from installation to that time).

[0118] Instead of the average of the log of the temperature gradient values, it is also possible to use any other statistical function such as the median and standard deviation.

[0119] Another example would be the case of equipment 2 such as a metal boiler tank, which is sensitive to and exposed to the temperature of the contained fluid, and is particularly subject to temperature cycles (i.e., high amplitude temperature fluctuations, e.g., between 200°C and 80°C, during the course of heating or cooling), and is sensitive to these cycles. The cycles can then be counted (e.g., 20 temperature cycles from installation to a certain point in time), and this count of similar cycles can be correlated with the average amplitude of the cycles in the log for that tank (e.g., average cycle amplitude of 150°C over 20 cycles).

[0120] Instead of the mean of the log of the amplitude, any other statistical function can be used, such as the median and the standard deviation. When the invention chooses to characterize the log of the condition 110 by the log of a function of a physical quantity representative of said condition 10, these functions can also include, without limitation, a count of said at least one variation.

[0121] This is especially the case if the equipment is sensitive to shutdown or start-up transients. A count of such transients is useful for characterizing the usage log 11. Similarly, if the equipment 2 continues to age while it is stopped, a count of periods of stoppage is also useful for characterizing the usage log. Similarly, in order to characterize the correlations 14, and in particular the condition 13 of the equipment 2, at a point in time as specifically as possible for the equipment 2 in question, measured physical quantities or functions of these measured physical quantities are also determined that characterize the material condition 13 of the equipment 2 at that point in time.

[0122] Consider for example the case of equipment 2 such as a centrifugal pump. The state of the pump can be instantaneously characterized by an instantaneous flow rate and an instantaneous discharge pressure. In this step, by technical analysis of the equipment 2, the present invention has determined the critical manufacturing and maintenance tasks 90 specific to the equipment 2 in order to best characterize the manufacturing and maintenance log 9 of the equipment 2. The usage conditions 110 specific to the equipment 2, as well as the physical quantities or functions of physical quantities that best characterize the conditions 110, as well as the usage log 11 of the equipment 2, have likewise been determined.

[0123] The material indicator 120 that best characterizes the condition 13 of the equipment 2 at any point in time, as well as the physical quantity or physical quantity function that best characterizes that material indicator 120 at that point in time, have likewise been determined.

[0124] This technical analysis of the device 2 therefore makes it possible to determine the correlation 14 in a form specific to the device 2. As a result, the determined correlation 14 is in fact specific to the physical phenomenon hosted by each device 2 of the series. Thus, according to one embodiment, in the correlation 14, the function of the measured physical quantity of the use condition 110 comprises a calculation of the time of existence of the measured physical quantity within at least one range of values ​​and / or a calculation representing at least one variation of the measured physical quantity and / or a count of said at least one variation.

[0125] Prior to this, in correlation 14, the invention chooses to characterize the manufacturing and maintenance log 9 of the equipment 2 (up to the last preemptive maintenance operation 10) by enumerating, for at least one (preferably each) critical manufacturing and maintenance task 90, the values ​​successively adopted during the manufacturing and various successive preemptive maintenance operations 10 in the life of the equipment 2 up to the last preemptive maintenance operation 10. Thus, the log of each critical task 90 is characterized by a list of values. Consider, for example, the aforementioned case of an equipment 2 such as a pump, equipped with bearings, and corresponding to the placement or replacement of the pump bearings and the critical tasks 90 of the nature for manufacturing and maintaining the pump. In this example, the last preemptive maintenance 10 (the one preceding the maintenance 6 to be performed) is considered to be the seventh (rank k=7). The present invention characterizes this log of critical manufacturing and maintenance tasks 90 by the list [A,0,0,B,0,0,B,0], indicating the placement of a type A bearing in manufacturing operation 4, the replacement of the bearing with a new type B bearing in the third maintenance operation, the replacement of the bearing with a new type B bearing in the sixth maintenance operation, and the fact that no maintenance work was performed on the bearing in the other maintenance operations.

[0126] The production and maintenance log 9 of the equipment 2 is therefore a matrix which groups together, for each critical task 90, the corresponding list of values. Moreover, with the aim of reducing the matrix characterizing the production and maintenance log 9 (up to the last predecessor maintenance operation 10) to the necessary and sufficient information, an amount of information independent of the rank of the last predecessor maintenance operation 10, the invention employs the following principle, referred to herein as the "persistent parameter principle". According to one embodiment, after reducing the production and maintenance log 9 in the correlation 14 to a log of critical tasks 90 in the form of at least one list of consecutive values, in each list only the persistent values ​​are selected as the values ​​adopted in the last maintenance operation in which the task 90 in question was performed, and only the persistent values ​​are kept in the log of critical tasks 90. In other words, in the production and maintenance log 9, for each critical production and maintenance task 90, the permanent value of the task 90 in question is considered to be the value adopted for the task 90 in the last manufacturing or maintenance step in which the task 90 was performed (the manufacturing step or the pre-maintenance step 10); - Only the persistence value of the task is kept in the manufacturing and maintenance log 9.

[0127] In its final form, the production and maintenance log 9 is therefore reduced to a list of values, i.e. a list consisting of the persistence values ​​of the critical tasks 90 (or one persistence value per critical task 90). In so doing, the invention reduces the production and maintenance log 9 of the equipment 2 by characterising said production and maintenance log 9 only by the values ​​of the critical tasks 90 which determine the behaviour of the equipment 2 during operation after the last predecessor maintenance operation 10 in question. In effect, in the production and maintenance log 9: the values ​​characterizing each manufacturing and maintenance task 90 carried out in the course of the last predecessor maintenance operation 10 overwrite the series of values ​​characterizing the log of that task 90 (from the manufacturing process 4 to the maintenance preceding the last predecessor maintenance operation 10); - for each task 90 not performed in the last preceding maintenance operation 10 from the manufacturing process 4 and from the preceding maintenance operation 10 preceding the last preceding maintenance operation 10, only the value characterizing the last execution of the task 90 is retained.

[0128] For example, consider the case of an equipment 2, such as a pump, for which the only two critical manufacturing and maintenance tasks 90 are a "place or replace pump bearings" task and a "place or replace pump impeller" task.

[0129] If the log of the "Pump bearing placement or replacement" task after the seventh predecessor maintenance operation 10 is characterized by the list [A,0,0,B,0,0,B,0], then the persistence value of the task 90 in question is "B", i.e., the value of the task 90 adopted during the last maintenance operation in which the task 90 in question was performed (in this example, during the sixth maintenance operation).

[0130] If the log of the "Pump Impeller Placement or Replacement" task after the seventh predecessor maintenance operation 10 is characterized by the list [C, 0, A, 0, 0, B, 0, C], then the persistence value of the task 90 in question is "C", i.e. the value of the task 90 adopted during the last maintenance operation in which the task 90 in question was performed (in this example, during the seventh maintenance).

[0131] Next, the pump manufacturing and maintenance log 9 after the seventh maintenance operation is written to [B;C], where the first term refers to the persistence value of the "arrange or replace pump bearings" task and the second term refers to the persistence value of the "arrange or replace pump impeller" task.

[0132] In other words, the present invention reduces the production and maintenance log 9 of the equipment 2 (up to and including the last leading maintenance operation 10) to a configuration in which the equipment 2 waits after the last leading maintenance operation 10. In doing so, the present invention characterizes the production and maintenance log 9 (up to the last leading maintenance operation 10) by a vector whose size is independent of the rank of said last leading maintenance operation 10. It is noted that the present invention can take into account the timing of the recorded values, for example in the form of a timestamp of the data.

[0133] In the characterization of the manufacturing and maintenance log 9, this timing may result in a timer 91 being added to each critical permanence value of a task 90, said timer 91 characterizing the duration since said task 90 was performed on the equipment 2. In this step, the correlations 14 previously determined by technical analysis of the equipment 2 in a form specific to the equipment 2 (and therefore specific to any standard equipment of the series) are then determined in a format suitable for computer processing, in particular by machine learning algorithms.

[0134] Once a technical analysis of the device 2 has been performed, series specific correlations 14 are identified. The following are then determined: - in the first section of correlation 14, the critical manufacturing and maintenance tasks 90 that best characterize the manufacturing and maintenance logs 9 of the equipment 2; in the second term of the correlation 14, the physical quantity or function of physical quantities that best characterize the condition 110 of the device 2 and the usage log 11, In the third term of the correlation 14, the physical quantity or function of physical quantities that best characterises the material indicator 120 and the log 12 of the condition 13 of the equipment 2.

[0135] The critical tasks 90, physical quantities or functions of physical quantities identified in the correlation 14 represent the raw data (measured and logged for each instrument 2 of the series) that are first extracted from the logs 9, 11, 12 of the instruments 2. The invention therefore foresees examining this accessible information coming from all instruments 2 of the series in question. For all instruments 2 of the series, the data associated with these tasks 90, as well as the data associated with the physical quantities or functions of physical quantities related to the conditions of use 110 and the material indicators 120 identified in the correlation 14, constitute a triplet 140 of data, which are recovered and extracted to obtain a raw data set 150. For each instrument 2 of the series and for each time point in their respective logs 11, the three terms of the triplet 140 related to that time point are: the part of the manufacturing and maintenance log 9 of the manufacturing and maintenance set 95 of the device 2, i.e. the part up to the last preceding maintenance 10 preceding the point in time, - a usage set 115 of the device 2, i.e. a part of the usage log 11 up to that point in time, - the state 13 of said equipment 2 at that time. As functions of the physical quantities identified in the correlations 14, they indicate a transformation 141 that is secondly applied to each of the triplets 140 of the raw data set 150 in order to constitute the final data set 15 that is taken into account for modelling the correlations 14 via the virtual model 16. Considering for example the case of an equipment 2 such as a pump, the terms of the correlations 14 specific to a series of pumps in question can be characterised as follows: - Manufacturing and maintenance log 9 is characterized by critical tasks 90 related to pump bearings and impellers, - the usage log 11 up to the time point is characterized by a first time integral of the temperature over the period between the installation and the time point (INT1(t)), a second time integral of the suction pressure over the period included between the installation and the time point (INT2(t)), and an average value of the temperature gradient (VMG(t)) over the period between the installation and the time point, and is associated with the number of sudden temperature transients (NTB(t)) over the same period, - A time state 13 is characterized by the time values ​​of the flow rate and the discharge pressure.

[0136] Data extracted from logs 9, 11, and 12 for each pump in the series are: - the type of bearing and the type of impeller currently installed, extracted from the manufacturing and maintenance log 9, for the first item of correlation 14, - for the second term of the correlation 14, the values ​​of the physical quantities at the time under consideration, i.e. temperature, suction pressure, extracted from the usage log 11, - for the third term of the correlation 14, it is the values ​​at the considered time of the physical quantities, i.e. the discharge pressure and the flow rate, extracted from the condition log 12. Once extracted, these data are, for each pump and for each time point (t) of their respective usage log 11, calculated at that time point, i.e. - it makes it possible to construct a manufacturing and maintenance set 95 of the pump, i.e. three terms of a triplet 140 relating to a part of the manufacturing and maintenance log 9 spanning the period between the manufacturing process 4 and the last preceding maintenance 10 preceding the instant (t) in question, which is in this example a log of tasks 90 (related to bearings and to the impeller of the pump) to be performed successively during the duration in question, or alternatively a matrix of two lists, one related to bearings and the other to the impeller of the pump, each list containing successive values ​​characterizing a critical task 90 performed during the period in question (a given type of bearing, a given type of impeller), a pump usage set 115, i.e. in this example the temperature and suction pressure values ​​recorded from installation up to the time (t), - List the state 13 of the pump at the time (t), i.e. in this example the flow rate and discharge pressure at the time (t).

[0137] Thus, as many triplets 140 are obtained as there are pumps of the series and time considered, and all these triplets 140 form the raw data set 150. Secondly, for each pump and each triplet 140 of the series associated with a time, the manufacturing and maintenance set 95 of the pump up to that time is transformed into the form of a vector associated with that time and indicating the permanence value of each critical task 90. ​​A corresponding vector is then formed, for each time considered, in which each term of the [type of bearing in position at time; type of impeller in position at time] vector is associated with its own timer 91. Similarly, for each pump and each triplet 140 of the series associated with a time, the usage set 115 of the pump up to that time is transformed. The functions of the physical quantities identified in the correlation 14 are, in this example, the integrals in time of temperature and suction pressure (INT1(t), INT2(t)), so that the mean value of the temperature gradient (VMG(t)) associated with the number of sudden temperature transients (NTB(t)), the raw data extracted from the usage log 11, temperature and suction pressure, are transformed in the transformation 141 applied to - Calculate the time integrals of temperature and suction pressure over the period between installation and time; - Calculate the average value of the temperature gradient over the same duration and the number of sudden temperature transients over the same duration.

[0138] Thus, for each pump in the series and for each time instant, the set of pump usages 115 between installation and time instant is converted into the form of a digital vector [INT1(t), INT2(t), VMG(t), NTB(t)] associated with the time instant (t). Then, for each pump in the series and for each time instant, the status 13 of the pump at that time instant is shown in the form of a digital vector associated with that time instant [flow rate, discharge pressure]. Finally, each of the triplets 140 (associated with the various pumps (k) belonging to the series and the various times (t) of their respective usage logs 11) is thus converted into the form of a vector [Mj(t)(k)]; Vu(t)(k); Status(t)(k)]. Mj(t)(k) is a vector representing the production and maintenance set 95 of pump(k) up to a point in time in the form [bearing type; wheel type], each term of the vector being associated with its timer 91. Vu(t)(k) is a vector representing the set of usages 115 of pump(k) between installation and time in the form [INT1(t), INT2(t), VMG(t), NTB(t)]. Status(t)(k) is a vector that represents the status 13 of pump (k) at time in the form of [flow rate, discharge pressure].

[0139] The raw data set 150 is thus transformed into the form of a final data set 15, i.e. a set of vectors [Mj(t)(k);Vu(t)(k);Status(t)(k)] for any pump (k) belonging to the series and for any time (t) of their respective usage logs 11. In short, the system 1 foresees recovering and processing only the data that it considers relevant, identified in the correlations 14. Optionally, the system 1 can recover all data related to the logs 9, 11, 12 of the equipment 2, in order to extract and then process only the relevant ones, as described above. Thus, the system 1 transforms the raw data set 150 into a data set 15 associated with the correlations 14 specific to the series of interest. The system 1 has this data set 15 as input for the machine learning project 142, and then at least one virtual model 16 is trained based on the data set 15. Like the correlations 14, the obtained virtual model 16 is specific to the equipment 2 of the series of interest. It further has the accuracy of an empirical model (data-driven model).

[0140] 5 and 6 illustrate the process of collecting data and training a model 16. In Fig. 5, logs 9, 11, 12 of equipment 2 and pending maintenance 6 for equipment 2 associated with a given point in time (t) during a time period 5 are considered.

[0141] FIG. 5 shows three terms of a triplet 140 linked by correlations 14: a production and maintenance set 95, a usage set 115 of the equipment, and a state 13 of the equipment at a given time (t); a production and maintenance set 95 including the production and maintenance decisions of rank 0 to rank k of the last maintenance preceding the time point (t), - a usage set 115 including the usage sublog 111 for the period 5 up to the point in time (t), - the state 13 of the device 2 at that time (t). Figure 15 also shows the extraction and processing of data relating to the device 2, in particular - extracting data associated with the production and maintenance set 95, the usage set 115, and the state 13 at a given time (t); - construction of a triplet 140 (set 95; set 115; state 13 at time t), as seen in FIG. 5; - the formation of as many similar triplets as there are time points (t) recorded during period 5, - injection of the triplet into the raw data set 150, as seen in FIG. 5; the composition of the raw data set 150 for each of the devices of the same type as device 2 in facility 3;

[0142] Figure 6 shows - recovering by transmission raw data sets 150 extracted from a number of facilities 3 operating the same series of equipment as the equipment 2, with the aim of merging them, and subsequently applying a transformation 141 to these data to form a data set 15; - presenting said data set 15 as input for a machine learning project 142, and training, on the basis of said data set 15, a virtual model 16 for modeling the correlations 14. By modelling the behaviour of an equipment 2 (or the behaviour of a series of equipment) in this way, the invention correlates, on the scale of the series, the observed behaviour (i.e. state 13) of each equipment 2 with the manufacturing and maintenance log 9 and the usage log 11 of said equipment 2. The invention is therefore a form of using feedback, which derives its remarkable properties from the relevance in a functional sense of the modelled correlations 14, in particular since they are specific to the physical phenomena hosted by each equipment 2 of the series.

[0143] According to one embodiment, the training of the model 16 belongs to the field of artificial intelligence and may be machine learning. In other words, the invention foresees the use of computer techniques, including artificial intelligence, in particular in the field of machine learning, based on mathematical and statistical approaches to give computers the ability to learn from data, i.e. to improve their performance in solving tasks without being explicitly programmed for each of them.

[0144] In particular, the monitoring system 1 comprises a model 16, which is predicted as a virtual, resulting from such learning. Furthermore, the training of the model 16 can include any type of machine learning, i.e., without limitation, supervised, semi-supervised, unsupervised, reinforced, or transfer classification. Furthermore, the machine learning can implement any category of training method, i.e., without limitation, the method can combine other methods such as neural networks (including deep learning methods), k-nearest neighbors ("KNN"), genetic algorithms, genetic programming, or Bayesian networks, support vector machines (SVM), Q-learning, decision trees, statistical methods, logistic regression, linear discriminant analysis, among others.

[0145] Preferably, the training of the model 16 involves a supervised, regression-based, deterministic machine learning problem (the model 16 preferably determines a vector of quantitative, continuous data from the set of labeled data 15, the data vector being selected by the present invention to integrate into the intended state 130 of the equipment 2).

[0146] According to other embodiments, the model 16 may be of any type.

[0147] The present invention, in the context of a machine learning project 142, envisages updating the virtual model 16 taking into account new production, maintenance, usage and condition data generated since the first training of the model 16. To do this, according to one embodiment, the above operations are repeated periodically, i.e. - the recovery of new data from at least one manufacturer, maintenance technician and / or operator is repeated. Preferably, new data relating to all devices 2 of the series are recovered from a set of operators who implement the devices of the series, - Extraction and transformation 141 of the new data is then performed to obtain a completed dataset 15, after which the training of the model 16 is updated on the basis of the completed dataset 15.

[0148] It should be noted that this updating is carried out periodically, the periodicity of which is determined by the designers, manufacturers, maintenance engineers and operators, the latter or according to the criteria considered most relevant from a "data science" point of view.

[0149] For example, the periodicity criterion may correspond to a ratio of 10% of the period over which the raw data set 150 was compiled in the first training context. If the data was compiled over a period of 10 years, the periodicity for updating the virtual model 16 may correspond to 12 months.

[0150] Thus, the present invention allows for the use of systematic feedback in a continuous or quasi-continuous manner.

[0151] According to the present invention, once the correlations 14 have been determined and the training of the model 16 has been performed, the present invention applies the model 16 to the device 2 .

[0152] To do this, advantageously, values ​​are submitted to said model 16 during a target maintenance 6 of said equipment 2. These values ​​correspond to at least one of the tasks 90 of the production and maintenance log 9 and to at least one of the usage conditions 110 of the usage log 11. Furthermore, values ​​correspond to at least one of the tasks 90 of the target maintenance 6 and to said planned usage conditions 110 of a scenario 8.

[0153] In effect, the model 16 learns to correlate the state 13 of the equipment 2 at a point in time with the production and maintenance set 95 (i.e., up to the last advance maintenance 10 preceding that point in time) and the usage set 115 up to that point in time.

[0154] Thus, the model 16 can correlate the planned state 130 of the equipment 2 at a point in time with the revised production and maintenance log 9 of the maintenance 6 tasks 90 to be performed, and the usage log 11 up to that point in time followed by the partial scenario 80.

[0155] In applying the model, the following input data is used: the overall manufacturing and maintenance log 9 of the equipment 2 (up to the last preceding maintenance 10 preceding the maintenance 6 to be performed), modified with a task 90 relating to the maintenance decision to be considered for the maintenance 6 to be performed; - a total usage log consisting of the usage log 11 of the equipment 2 (up to the day before the maintenance 6 to be performed) followed by partial scenarios 80 of the target (from the start of the planned operating period 70 to the point considered in the course of said planned period 70) is submitted to the model 16.

[0156] Thus, the present invention uses the model 16 as a customized simulator for the device 2 .

[0157] Indeed, the model 16 makes it possible to simulate the planned state 130 of the equipment 2 at a certain point in time for a given maintenance decision regarding the task 90 of the maintenance 6 to be performed on the equipment 2 in question, and for the usage of the equipment 2 that follows in the context of the scenario 8 up to that point in time. Moreover, this simulation is customized to the equipment 2 in question, since the production and maintenance log 9 up to the day before the maintenance 6 to be performed, and the usage log 11 up to the day before the maintenance 6 to be performed, taken into account to perform this simulation, are specific to the equipment 2, i.e. specific to the equipment 2 in question.

[0158] As can be seen in FIG. 1 , the application of the monitoring system 1 to a piece of equipment 2 pending a scheduled maintenance 6 (of rank j) makes it possible to determine a planned state 130 of said equipment 2 for a future point in time corresponding to the day before a scheduled maintenance 7 (of rank j+1), and to obtain data relating to said equipment 2, - The device's manufacturing and maintenance log 9 and usage log 11, - taking into account the maintenance task 90 regarding the maintenance to be performed 6 and the use scenario 8.

[0159] In particular, FIG. 1 shows a model 16 that receives the aforementioned data as input and produces as output a planned state 130 of the equipment 2 at the end of a planned period 70 .

[0160] Furthermore, when these values ​​are submitted, the model 16 generates a planned state 130 of the equipment 2 after the maintenance 6 is performed.

[0161] In other words, from the data submitted as input, the model 16 is able to predict a planned state 130 of the equipment 2 for the time considered, i.e. the values ​​for the time considered for each of the material indicators 120 of this planned state 130. The planned state 130 is then compared with the minimum states 17 identified as necessary for the operation of the equipment 2.

[0162] In other words, for each of the material indicators 120 of this planned state 130, the value predicted by the model 16 is compared with the minimum value required for the operation of the equipment 2, e.g. the minimum value of the performance of the equipment 2 (or, depending on the situation, with the maximum value required for the operation of the equipment 2, e.g. the maximum value of the wear of the equipment 2).

[0163] If the planned state 130 thus simulated for the time point under consideration complies with the operating criteria (i.e., if the values ​​predicted by the model 16 of each material indicator 120 of the planned state 130 are greater (or less) than the minimum (or maximum) values ​​required for operation of the equipment 2), then the predicted planned state 130 corresponds to the correct operating state of the equipment 2.

[0164] For example, the case of a device 2 such as a pump is considered, whose state 13 can be considered at a time (t) as characterized by two material indicators 120, namely, flow rate (Q(t)) and discharge pressure (Pref(t)), respectively, relative to the minimum required for the operation of the pump, Q min and Pref min It is.

[0165] Therefore, the minimum state of the pump is the set of values ​​[Q min ,Pref min Therefore, the operation criterion for the pump at time (t) is that the planned state 130 is greater than the minimum state 17, i.e., the flow rate Q(t) and the discharge pressure Pref(t) satisfy the two conditions Q(t)>Q min and Pref(t)>Pref min The purpose is to satisfy the following.

[0166] To identify whether the planned state 130 of the pump corresponds to the correct operating state of the pump as predicted by the model 16 for the time instant considered, the state 130 is then compared to the minimum state 17, i.e., the predicted values ​​of the flow rate Q(t) and discharge pressure Pref(t) material indicators 120 for the time instant are compared to their respective minimum values ​​Q(t) and Pref(t) min and Pref min is compared to.

[0167] FIG. 7 shows a prediction of this planned state 130 and a comparison with the minimum state 17 for this example.

[0168] FIG. 7 repeats the previous example for an equipment 2, such as a pump, whose state 13, at a time (t), can be characterized by two material indicators 120, namely, flow rate (Q(t)) and discharge pressure (Pref(t)), where Q min and Pref min and , respectively, are the values ​​of the material indicator 120 in the minimum state 17, which is the minimum value required for the pump to operate.

[0169] FIG. 7 shows the progression of the planned state 130 of the pump (ie, the successive values ​​of the material indicator 120Q and Pref) calculated by the model 16 over the course of a planned period 70.

[0170] Since the curve of the expected state 130 decreases over time due to aging, FIG. 7 also shows a comparison of the state 130 thus predicted at a time point (t) with the minimum state 17. To identify whether the expected state 130 of the pump corresponds to the correct operating state of the pump, the expected state 130 is then compared with the minimum state 17, i.e., the predicted values ​​of the flow rate (Q(t)) and discharge pressure (Pref(t)) material indicators 120 for that time point are compared with their respective minimum values ​​Q(t) and P(t) of the material indicators 120. min and Pref min is compared to.

[0171] According to an embodiment, for a given maintenance decision for a maintenance to be performed 6 (decision for a task 90 to be performed on equipment 2) followed by a given scenario 8, it is determined whether the maintenance decision is sufficient, i.e. whether it gives the equipment 2, in the context of the current scenario 8, a sufficient probability of zero failures in operation until the end of the scheduled period 70 or until the day before the scheduled maintenance 7. To do this, according to an embodiment, at least one variation of at least one of the values ​​of the tasks 90 of the maintenance to be performed 6 is performed. The value of the variation is then introduced when the values ​​are submitted to the model 16.

[0172] Among all the variations, at least one sufficient determination of the maintenance to be performed is selected for the scheduled state 130 of the equipment 2 to be equal to or greater than the minimum state 17 at the time of the scheduled maintenance 7 .

[0173] In other words, if the planned maintenance 7 corresponds to the next maintenance step following the to-be-performed maintenance 6, then a set of possible maintenance decisions (relating to the tasks 90) is considered for the to-be-performed maintenance 6, with the intention of determining among the latter a sufficient maintenance decision. For example, two maintenance tasks 90 are identified as critical: Replace pump bearings (types A and B are used as possible bearing types), Consider an appliance 2, such as a pump, for which the pump impeller is replaced (using types A, B, and C as possible impeller types).

[0174] The possible maintenance decisions for the maintenance 6 of the implementation object are 12 combinations of decisions formed from the following: - Three possible decisions regarding the bearing: do nothing (leave the bearing in place) or replace the bearing with a type A or type B bearing. - Four possible decisions regarding the impeller: do nothing (leave the impeller in place) or replace the impeller with a Type A, Type B, or Type C impeller.

[0175] Among these 12 possible maintenance decisions, it is intended to determine a sufficient maintenance decision. Furthermore, among these possible decisions of the maintenance 6 to be performed, if a maintenance decision is characterized by the vector [0;A] (i.e., the bearing is not replaced and the impeller is replaced with a type A impeller) and another maintenance decision is characterized by the vector [0;0] (i.e., the bearing is not replaced and the impeller is not replaced), then these two maintenance decisions for the maintenance 6 to be performed will change with respect to the impeller (in extenso, the change in the value of task 90 of the maintenance 6 to be performed will relate to the change in the "replace" and "not replace" values ​​of the impeller). For each possible maintenance decision for the target maintenance 6, the aforementioned customized simulation is implemented by the model 16 to determine a planned state 130 of the equipment 2 at the end of the planned period 70, i.e. on the day before the planned maintenance 7 (taking into account the usage of the equipment 2 following the target maintenance 6 in the context of the scenario 8 up to the time point corresponding to the day before the target maintenance 7, taking into account the production and maintenance log 9 of the equipment 2 up to the day before the target maintenance 6, and taking into account the usage log 11 of the equipment 2 up to the day before the target maintenance 6). If the planned state 130 (thus simulated by the model 16 for the time point corresponding to the end of the planned period 70) meets the operation criteria of the equipment 2 compared to the minimum state 17 required for the operation of the equipment 2, the maintenance decision is considered sufficient to give the equipment 2 a sufficient probability of zero failures during operation in the context of the scenario 8 up to the end of the planned period 70 (referred to as a "sufficient maintenance decision").

[0176] For example, consider the case of a device 2 such as a pump. Its state 13 has two material indicators 120 in state 13, namely (Q min ) and (Pref min The maintenance decision can be made based on the fact that the planned state 130 (for a time corresponding to the end of the planned period 70) satisfies two conditions on the pump operation criteria, i.e., the material indicator 120 for a time corresponding to the end of the planned period 70: (Q>Q min) and (Pref>Pref min ) is considered sufficient. The simulation is repeated for each of the possible maintenance decisions (for tasks 90) for the maintenance 6 to be performed. Thus, each planned state 130 simulated by the model 16 for a time point corresponding to the end of the planned period 70 makes it possible to sort the possible maintenance decisions for the maintenance 6 to be performed between sufficient and insufficient decisions.

[0177] Thus, among all variations in the values ​​of at least one of the tasks 90 of the maintenance 6 to be performed, at least one sufficient determination of the maintenance 6 to be performed is selected for a minimum state 17 or greater for the planned state 130 of the equipment 2 at the time of the planned maintenance 7.

[0178] According to one embodiment, for a given maintenance decision (related to task 90 of the maintenance 6 to be performed) and for each planned usage condition 110 of scenario 8 of equipment 2 over planned period 70, a maximum possible limit 18 (i.e. a limit that is compatible with uninterrupted operation of equipment 2 without risk of failure) is determined that remains compatible with zero failures of equipment 2 in the context of scenario 8 until the middle and end of planned period 70 or until the day before the next planned maintenance 7 following the maintenance 6 to be performed.

[0179] To do this, for a given maintenance decision, the value of at least one planned use condition 110 of a scenario 8 is modified. The value of the modification as well as the value of the maintenance decision are introduced when the value is submitted to the model 16. A limit is then calculated for at least one of the planned conditions 110 for a planned state 130 of the equipment 2 equal to a minimum state 17 at the time of the planned maintenance 7. In other words, a scenario 8 of the equipment 2 over the planned period 70 is characterized by at least one controllable planned use condition 110, the value of which ("remains free") remains free to change, while the other possible planned use conditions 110 (controllable and / or non-controllable) are set to the respective values ​​of the scenario 8. The model 16 is then used to solve, by an iterative method, the equations in which the planned state 130 of the equipment 2 at the end of the planned period 70 corresponds to the minimum state 17 for the freed controllable planned use condition 110.

[0180] For a given maintenance decision (relating to the maintenance 6 task 90 to be performed) and the target controllable planned use conditions 110, a possible limit (preferably a maximum limit 18) is thus determined that remains compatible with zero failures of the equipment 2 until the end of the planned period 70 (i.e. a limit that is compatible with uninterrupted operation of the equipment 2 without risk of failure) (the other planned use conditions 110 remain set to their respective values ​​in scenario 8) (the other planned use conditions 110 remain set to their respective values ​​in scenario 8).

[0181] For example, if the facility 3 is a truck and the monitored equipment 2 is the engine of said truck, then the scenario 8 between the maintenance to be performed 6 and the next scheduled maintenance 7 is considered, and the following scheduled usage conditions 110 are considered: - controllable intended conditions of use 110: average load to be transported (2 tons = 2T), distance to be traveled (100,000 km = 100,000 km), average speed (90 kilometers per hour = 90 km / h), - Uncontrollable intended conditions of use 110: characterized by the ambient air temperature (20 degrees Celsius = 20°C), the average gradient of the route to be used (5 percent = 5%).

[0182] In this scenario 8, the intention is to determine the value of the maximum limit 18 of the average cargo transported that is compatible with zero failures until the scheduled maintenance 7 following the maintenance to be performed 6, while the other planned usage conditions 110 remain set to their respective target values ​​of scenario 8.

[0183] The use of the model 16 to solve, by iteration, for the average load transported, the equation for which the planned state 130 of the equipment 2 at the end of the planned period 70 corresponds to the minimum state 17 gives, by way of example, a maximum limit 18 of 2.5T (for a target value of the average load transported of 2T in scenario 8). In FIG. 8, the aforementioned example of an equipment 2 such as a truck engine is considered, with the operating conditions 110 being the load P to be transported, the distance traveled d, the average speed V, the ambient air temperature T°, and the average gradient A of the route used. In FIG. 8, the planned state 130 is defined by a material indicator 120 corresponding to the average compression pressure U of the cylinders, and the value U of the minimum state 17 is taken as the minimum value necessary for the correct operation of the equipment 2. min 8 shows the calculation of the maximum limit 18 (Plim) of the controllable usage conditions 110P for the equipment 2 associated with its fixed logs 9 and 11 for a given maintenance decision regarding the maintenance 6 to be performed for a usage scenario 8 over a planned period 70 defined by the target average values ​​(Pc, dc, Vc, T°, Ac) of the respective controllable (P, d, V) and uncontrollable (T°, A) usage conditions 110 (recall from the previous example that Pc=2T; dc=100,000 km; Vc=90 km / h; T°c=20° C.; Ac=5%).

[0184] For the other four operating conditions 110 (d, V, T°, A) set to the respective target average values ​​of scenario 8 (dc=100,000 km, Vc=90 km / h, T°c=20° C., Ac=5%), the maximum limit 18 of the controllable operating conditions 110 (P) (determined as Plim=2.5T in this example) is iteratively determined by the model 16.

[0185] For this maximum limit 18, Plim, the scheduled state 130 at the end of the scheduled period 70 corresponds to the required minimum state 17, i.e., the material condition indicator 120 is at the required minimum value (U=U min )

[0186] Similarly, FIG. 9 shows the calculation of the maximum limit 18 (dlim) for the controllable operating condition 110d. For the other four operating conditions 110 (P, V, T°, A), the respective target average values ​​of scenario 8 (Pc=2T; Vc=90km / h; T°c=20°C; Ac=5%) are set, and the maximum limit 18 for the controllable operating condition 110(d) is iteratively determined by the model 16 (in this example, determined as dlim=110,000km). For this maximum limit 18 (dlim), the planned state 130 at the end of the planned period 70 corresponds to the required minimum state 17, i.e. the material condition indicator 120 is set to the required minimum value (U=U min) for each of the other controllable planned use conditions 110 of the scenario 8. The same is done for a given maintenance decision (related to a task 90 of the maintenance 6 to be performed) to determine its possible maximum limit 18 (i.e. a limit that is compatible with uninterrupted operation of the equipment 2 without risk of failure) that remains compatible with zero failure of the equipment 2 in the context of the scenario 8 during and until the end of the planned period 70, i.e. until the day before the next planned maintenance 7 following the maintenance 6 to be performed. Once determined in this way, said maximum limit 18 defines, for a given maintenance decision, the bounds of a usage range that is compatible with zero failure of the equipment 2 in the context of the scenario 8 during and until the end of the planned period 70. It is therefore known that as long as the equipment 2 is operated, zero failure of the equipment 2 is possible and the given controllable planned use condition 110 is kept below the value of its maximum limit 18 thus determined, while the other planned use conditions 110 remain below their respective target values ​​of the scenario 8. It is then possible to graphically represent by means of a web mapping diagram the target values ​​of the planned conditions of use 110 of the scenario 8 as well as the previously calculated values ​​of the maximum limits 18 for each of the controllable planned conditions of use 110, each axis being assigned to a planned condition 110. A polygon is thus obtained with the same number of vertices as there are planned conditions 110.

[0187] Thus, a first polygon 180, called "target polygon 180" (representing the target values ​​of each of the planned conditions of use 110 of scenario 8) can be superimposed with a second polygon 181, called "limit polygon 181" (representing the values ​​of the maximum limits 18 of each of the controllable planned conditions of use 110 and the target values ​​of scenario 8 for each of the non-controllable planned conditions of use 110), the target polygon 180 being surrounded by the limit polygon 181. In FIG. 10, the aforementioned example of an equipment 2, such as a truck engine, is considered, the conditions of use 110 being the load P to be transported, the distance d to be traveled, the average speed V, the ambient air temperature T°, and the average gradient A of the route used. In FIG. 10, the planned state 130 is defined by the material condition indicator 120, which corresponds to the average compression pressure U of the cylinders, and the value U of the minimum state 17 is taken as the minimum value required for the correct operation of the equipment 2.min 10 shows a target polygon 180 and a limit polygon 181 of the equipment 2 associated with its fixation logs 9 and 11 for a given maintenance decision regarding the maintenance 6 to be performed for a usage scenario 8 over a planned period 70 defined by target average values ​​(Pc, dc, Vc, Tc, Ac) for each of the controllable (P, d, V) and uncontrollable (T°, A) usage conditions 110 (recall from the previous example that Pc=2T; dc=100,000 km; Vc=90 km / h; T°c=20° C.; Ac=5%).

[0188] FIG. 10 shows a "target polygon 180" in solid lines along the corresponding axis, - the target mean values ​​(Pc,dc,Vc) of each of the controllable planned use conditions 110 (P,d,V) of scenario 8, represented by an open padlock; - showing the target mean values ​​(T°c,Ac) for each of the uncontrollable planned use conditions 110 (T°,A) of scenario 8, represented by closed padlocks.

[0189] FIG. 10 shows a "constraint polygon 181" in dotted lines along the corresponding axis, - the values ​​(Plim, dlim, Vlim) of the maximum limits 18 of each of the controllable planned conditions of use 110 (P, d, V) of scenario 8, represented by an open padlock; - showing the target mean values ​​(T°c,Ac) for each of the uncontrollable planned use conditions 110 (T°,A) of scenario 8, represented by closed padlocks.

[0190] According to one embodiment, the calculation of the maximum limits 18 of the planned usage conditions 110 of scenario 8 is repeated for each maintenance decision determined to be sufficient (i.e., a maintenance decision for a task 90 to be performed on the equipment 2 during the planned maintenance 6 that, in the context of that scenario 8, gives the equipment 2 a sufficient probability of zero failures during operation until the middle and end of the planned period 70 or by the day before the planned maintenance 7).

[0191] To do this, according to one embodiment, a selected value of the sufficient maintenance determination is introduced when values ​​are submitted to the model 16. The maximum limit 18 of the planned operating conditions 110 is calculated for this sufficient maintenance determination, for the planned state 130 of the equipment 2 at the time of the planned maintenance 7 to be equal to the minimum state 17.

[0192] When the maximum limit 18 of the controllable planned use conditions 110 has been calculated for each maintenance decision determined to be sufficient to implement maintenance 6 and for the scenario 8 considered, the target usage in the form of a target polygon 180 (i.e. the target value of the planned use conditions 110 of scenario 8) and the limit usage in the form of a limit polygon 181 (i.e. the value of the maximum limit 18 of the controllable planned use conditions 110 of scenario 8 and the target value of the uncontrollable planned use conditions 110) are graphically represented on a similar web mapping diagram by superimposing the two polygons.

[0193] According to one embodiment, a usage margin is determined for at least one of the planned usage conditions 110 of the scenario 8 as the deviation 182 between a corresponding value and a corresponding maximum limit 18 .

[0194] In fact, the present invention quantifies, for all controllable planned use conditions 110 of scenario 8 , the deviation 182 between their target values ​​and the values ​​of their respective maximum limits 18 .

[0195] In other words, the representation and overlay on the same web mapping diagram of the target polygon 180 of scenario 8 and the limit polygon 181 of the maximum limit 18 makes it possible to graphically represent, for each controllable planned use condition 110, the deviation 182 between the target value and the value of the maximum limit 18 of that controllable planned use condition 110.

[0196] For example, in the case of the truck engine mentioned above, the deviation 182 between the target value and the value of the maximum limit 18 is calculated for each controllable planned operating condition 110 as follows: - for the average load transported: deviation between the target value 2T and the maximum limit value 18 2.5T 182, - for the distance to be travelled: the deviation 182 between the target value of 100,000 km and the value of the maximum limit 18 of 110,000 km, - for the average speed: the deviation 182 between the target value 90 km / h and the value of the maximum limit 18 100 km / h is represented graphically.

[0197] 10 shows, for each controllable planned use condition 110, the deviation 182 between the target value and the value of the maximum limit 18 of said controllable planned use condition 110. Preferably, in order to compare the comparable deviations 182, the axes of the web mapping diagram on which the target polygon 180 of scenario 8 and the limit polygon 181 of the maximum limit 18 are shown and superimposed are normalized. In other words, for each controllable planned use condition 110, the target value as well as the value of its maximum limit 18 are represented on a scale and each value is normalized (i.e. represented by the ratio between it and the target value).

[0198] For example, in the case of the truck engine mentioned above, on the axis-normalized web mapping diagram, the deviation 182 between the target value and the maximum limit 18 value for each controllable planned operating condition 110 is: - for the average load to be transported, the target value 2T is represented by the ratio 1 (2T / 2T), while the value of the maximum limit 18 is represented by the ratio 1.25 (2.5T / 2T), i.e. a deviation of 25% between the two values; - for the distance to be travelled, the target value is expressed in the ratio 1 (100,000km / 100,000km) and the value of the maximum limit 18 is expressed in the ratio 1.1 (110,000km / 100,000km), i.e. the deviation between these two values ​​is 10%; - for the average speed: the target value can be shown in the graph to be represented by the ratio 1 ((90km / h) / (90km / h)), while the value of the maximum limit 18 is represented by the ratio 1.11 ((100km / h) / (90km / h)), i.e. the deviation between the two values ​​is 11%. These deviations 182 between the target value and the value of the maximum limit 18, normalized, therefore serve for comparison with each other. They make it possible to identify controllable planned conditions of use 110 in which the deviation 182 between the target value and the value of the maximum limit 18 is significant or acceptable, as well as controllable planned conditions 110 in which the deviation 182 is not, which amounts to identifying planned conditions of use 110 that allow a low or high margin before failure, whether acceptable or not.

[0199] For example, in the case of the truck engine mentioned above, the deviation 182 between the target value and the value of the maximum limit 18 is much larger for the average load transported (25%) than for the mileage and average speed traveled (10% and 11%, respectively). Therefore, the average load transported tolerates a larger deviation 182 from its target value than the mileage and average speed traveled (without, however, questioning zero failures). In other words, the average load has a higher pre-failure reliability than the mileage and average speed traveled, in order to benefit from zero failures until the end of the scheduled period 70. Therefore, the operator should pay more attention to observing the target values ​​of scenario 8 for the mileage and average speed traveled than the average load transported during the scheduled period 70. Therefore, the present invention defines the usage margin before failure (MUBF) as the average of the normalized deviations 182 between the target values ​​and the maximum limit 18 values ​​for each controllable planned usage condition 110, or the average deviation between the target polygon 180 of the scenario 8 and the limit polygon 181 of the maximum limit 18.

[0200] In the above example, the average of the three normalized deviations 182, 25%, 10% and 11%, is a margin of failure (MUBF) of approximately 15%. Therefore, for the maintenance decisions determined to be sufficient for the maintenance 6 to be performed and for the scenario 8, the maximum limit 18 of the controllable scheduled conditions 110 is calculated, and the suitability of the maintenance decision (with respect to the task 90 of the maintenance 6 predicted to be performed) is preferentially quantified in the scenario 8 and in the complete log of the equipment 2 by the margin of failure (MUBF) indicator associated with the maintenance decision.

[0201] For example, consider the case of equipment 2, such as a truck engine. Seven sufficient maintenance decisions are identified for the maintenance 6 to be performed, and a margin of failure (MUBF) is calculated for each of them. These decisions are: - a first determination (D1) which allows changing the diesel filter and obtaining a MUBF of 3%; - A second determination (D2) which allows replacing the injector and obtaining a MUBF of 7%; - A third determination (D3), which allows replacing the injectors and diesel filters and obtaining a MUBF of 9%; - The fourth decision (D4), which replaces the injection pump and allows to obtain a MUBF of 11%; - The fifth determination (D5), which allows replacing the injection pump and diesel filter and obtaining a MUBF of 13%; - The sixth determination (D6), which allows replacing the injection pump and injectors and obtaining a MUBF of 20%; - The seventh determination (D7) which allows replacing everything (injection pump, injectors and diesel filter) and obtaining a MUBF of 25%.

[0202] It is noted that, with regard to the equipment 2 and its logs 9, 11, decision D7 proves to be more suitable for scenario 8 than the other decisions, especially decision D1. Indeed, decision D1 makes it possible to obtain a MUBF of 3%, the mean value of the 3% deviation 182 for the planned conditions of use 110 being to be respected with respect to the target value of scenario 8, in order to benefit from zero failures until the end of the planned period 70.

[0203] Decision D7, for its part, requires that in order to allow a MUBF of 25% and to benefit from zero failures by the end of the planned period 70, the average deviation 182 for the planned conditions of use 110 that is much larger than 25% should be respected with respect to the target value of scenario 8.

[0204] The Margin of Use (MUBF) therefore makes it possible to quantify the suitability of each conservation decision for the same planned usage scenario8.

[0205] According to an embodiment, once the usage margin is obtained, the optimal maintenance decision is selected from among the sufficient maintenance decisions as the sufficient decision that allows to obtain an acceptable usage margin or as the sufficient decision that allows to obtain at least one of said acceptable deviations 182. In other words, for each of the decisions identified as sufficient to perform the maintenance 6 and for the scenario 8 under consideration, the maximum limit 18 of the controllable planned usage conditions 110 is calculated, the suitability of the sufficient maintenance decision (with respect to the task 90 for the maintenance 6 foreseen to be performed) for the scenario 8 is quantified by the usage margin before failure (MUBF), and the invention preferentially selects the optimal maintenance decision as the decision that allows the intended usage margin. Thus, the optimal decision with respect to the amount of the task 90 for the maintenance 6 to be performed is preferentially determined. In short, this optimal decision is customized for the complete log of the equipment 2 and is suitable for the scenario 8 (allowing zero failures until the end of the planned period 70 as well as the intended usage margin before failure).

[0206] For example, consider the aforementioned case of a truck engine where seven sufficient maintenance decisions were identified (D1-D7) for the maintenance 6 to be performed and a margin (MUBF) was calculated for each.

[0207] The operator determines that a 10% deviation 182 between the target value of scenario 8 and the value of the maximum limit 18 is - for the loads to be transported (i.e. the operator estimates that the average load to be transported between the scheduled maintenance 6 and the next scheduled maintenance 7 will not exceed the target of 2T by more than 10%), - the total number of miles to be traveled until the next maintenance (i.e., the operator estimates that the total number of miles to be traveled between the scheduled maintenance 6 and the next scheduled maintenance 7 will not exceed the target of 100,000 kilometers by more than 10%); - regarding the average speed (i.e. the operator estimates that the average speed between the maintenance to be performed 6 and the next scheduled maintenance 7 will not exceed the target of 90 km / h by more than 10%) can be considered sufficient.

[0208] Therefore, the sufficient maintenance decision for the to-be-performed maintenance 6, which is associated with a margin (MUBF) of more than 10% and minimizes the maintenance constraint, is decision D4 with a margin (MUBF) of 11%. In particular, decisions D5, D6 and D7 also seem to allow for a margin (MUBF) greater than 10%, but would entail unnecessary over-maintenance with a higher maintenance constraint in terms of the operator's intent. Therefore, for the operator, the optimal maintenance decision for the to-be-performed maintenance 6 is decision D4 (i.e., the decision that includes replacing the injection pump).

[0209] The present invention also contemplates selecting an optimal maintenance decision to be applied during the maintenance 6 to be performed as the decision that allows for an intended deviation 182 between the target value of the scenario 8 and the value of the maximum limit 18, and does so for one of the planned usage conditions 110.

[0210] For example, consider the aforementioned case of a truck engine where seven sufficient maintenance decisions (D1-D7) were identified for maintenance 6 to be performed, and the operator: - Estimate that the mileage traveled on the day before the next scheduled maintenance 7 cannot exceed the target of 100,000 kilometers in scenario 8, - Estimate that the average speed on the day before the next scheduled maintenance 7 will not be able to exceed the target of 90 km / h in scenario 8, - For the average load transported only, a deviation 182 of at least 10% between the target value (2T) and the value of the maximum limit 18 is intended.

[0211] In this example, for each of the seven sufficient maintenance decisions, the deviation 182 for the load to be transported between the target value in scenario 8 (2T) and the value of the maximum limit 18 calculated by the simulation of the model 16 is considered, - Decision D1 allows for a deviation of 5% for the load to be transported 182; - Decision D2 allows for a deviation of 11% for the load to be transported182; - Decision D3 allows for a deviation of 15% for the load to be transported, - Decision D4 allows for a deviation of 18% for the load to be transported, - Decision D5 allows for a deviation of 21% for the load to be transported182; - Decision D6 allows for a deviation of 33% for the load to be transported182; - Decision D7 allows a deviation 182 of 41% for the load to be transported.

[0212] Taking into account the operator's assumptions and intentions, the optimal maintenance decision for the maintenance 6 to be performed is in this case no longer decision D4 but decision D2 (with a deviation 182 of 11% between the target value for the load to be transported and the value of the maximum limit 18). In particular, decisions D3 to D7 also seem to result in a deviation 182 of more than 10%, but with respect to the operator's intentions, which would involve unnecessary over-maintenance with higher maintenance constraints. Therefore, for the operator, the optimal maintenance decision for the maintenance 6 to be performed is in this case decision D2 (i.e. the decision involving replacing the injector).

[0213] The present invention can also select the optimal maintenance decision to be applied during the maintenance 6 to be performed as the decision that allows for the intended average of deviations between the target values ​​of the scenario 8 and the values ​​of the maximum limits 18, and does so for the intended portion of the planned usage conditions 110.

[0214] For example, consider the aforementioned case of a truck engine where seven sufficient maintenance decisions (D1-D7) were identified for maintenance 6 to be performed: - The operator estimates that the average speed on the day before the next maintenance will not be able to exceed the target of 90 km / h, The operator intends an average deviation 182 of 10% (between the target values ​​of the scenario 8 and the values ​​of the maximum limit 18) for the loads transported and the distances traveled.

[0215] In this example, for each sufficient maintenance determination, the average of the deviations 182 between the target value of scenario 8 and the maximum limit 18 is considered for two planned use conditions 110 (cargo transported and mileage traveled) with maximum limits 18 calculated by simulation allowed by model 16.

[0216] In this example, - Decision D1 allows deviations of 5% and 2% for the loads transported and the distances traveled, respectively,182 or an average deviation of 3%; - Decision D2 allows deviations of 11% and 5%, or a mean deviation of 8%, - Decision D3 allows deviations of 15% and 6%,182 or a mean deviation of 10%; - Decision D4 allows deviations of 18% and 7%, or a mean deviation of 13%; - Decision D5 allows deviations of 21% and 8%,182 or a mean deviation of 15%; - Decision D6 allows deviations of 33% and 13%, or a mean deviation of 23%; It is further assumed that decision D7 tolerates deviations 182 of 41% and 16%, i.e. a mean deviation of 29%.

[0217] Taking into account the operator's assumptions and intentions, the maintenance decision for the maintenance 6 to be performed is in this case no longer decision D2 or decision D4 but decision D3 with a mean deviation of 10%. In particular, decisions D4-D7 also seem to tolerate a mean deviation of more than 10%, but with respect to the operator's intentions, to involve unnecessary over-maintenance with a higher maintenance constraint. Therefore, for the operator, the optimal maintenance decision for the maintenance 6 to be performed is in this case decision D3 (i.e. the decision involving replacing the injectors and diesel filters).

[0218] Thus, for a piece of equipment 2 that has been operated up to the maintenance 6 to be performed, and for a scenario 8 that is executed after said maintenance 6 to be performed, the simulation allowed by the model 16 makes it possible to determine: - An optimal maintenance decision among a set of sufficient maintenance decisions: this decision on the nature of the task 90 for the maintenance 6 to be performed is customized for the complete log of the equipment 2 and is suitable for the scenario 8 to allow zero failures until the end of the planned period 70 and to allow the intended usage margin before failure (MUBF). - the associated limit usage, i.e. the maximum possible usage bounded by the maximum limit 18 of the controllable planned usage conditions 110 (which maximum limit 18 is associated with the optimal maintenance decision) (meeting zero failures until the date of the next planned maintenance 7 following the maintenance 6 to be performed).

[0219] For example, consider the aforementioned case of a truck engine where seven sufficient maintenance decisions (D1-D7) have been identified for the maintenance 6 to be performed and the operator intends a mean usage margin before failure (MUBF) of 10% for a set of controllable planned usage conditions 110 (average load transported, mileage traveled until next maintenance, and average speed).

[0220] In this example (and considering the manufacturing and maintenance log 9 and usage log 11 of the truck engine, and scenario 8 over the course of the planned period 70), among the sufficient maintenance decisions for the maintenance 6 to be performed, the optimal maintenance decision is decision D4 (i.e., one that includes replacing the injection pump).

[0221] In order to benefit from zero engine failures by the end of the planned period 70, the limit usage observed during the course of the planned period 70 is the limit usage associated with this optimal maintenance decision (i.e., limited by the planned usage conditions 110 and the maximum limits 18 calculated for the maintenance decision D4: 2.3T with an 18% deviation 182 from the target value of 2T, 107,000km with a 7% deviation 182 from the target value of 100,000km, 97km / h with an 8% deviation 182 from the target value of 90km / h).

[0222] In other words, taking into account the mode for calculating the maximum limit 18, it is possible to guarantee zero failures of the truck engine until the end of the planned period 70 as long as the engine is operated with given controllable conditions 110 maintained below the value of the maximum limit 18 thus determined and other conditions 110 maintained below the respective target values ​​of the scenario 8.

[0223] The construction of a constraint polygon 181 from the maximum constraints 18 of the intended operating conditions 110 to quantify the mean deviation (i.e., utilization margin before failure (MUBF)) of a scenario 8 from a target polygon 180 is a conceptual tool envisaged by the present invention.

[0224] Indeed, the limiting polygon 181 and the margin (MUBF), which are always calculated according to the same method in each case of sufficient maintenance decisions 6 taken for the same target scenario 8, make it possible to evaluate the conformity of each maintenance decision with respect to the target scenario 8 (taking into account the manufacturing and maintenance logs 9 and the usage logs 11 of the equipment 2). The limiting polygon 181 and the margin (MUBF) therefore make it possible to compare the maintenance decisions between them.

[0225] Furthermore, the present invention foresees extrapolating all usage scenarios that meet the zero failures until next maintenance date.

[0226] Indeed, considering the mode for calculating the maximum limit 18, the limit polygon 181 and its maximum limit 18 limit the limit usage that complies with zero failures of the equipment 2 until the end of the planned period 70, in the sense that zero failures are possible only if the equipment 2 is operated in a given controllable planned usage condition 110 that is kept below the value of its maximum limit 18, but if the other planned usage conditions 110 are kept below their respective target values ​​of scenario 8. In particular, the limit polygon 181 does not allow to estimate whether the usage of the equipment 2 having two controllable planned usage conditions 110 that can exceed their respective target values ​​while observing their respective possible maximum limits 18 remains in compliance with the intended zero failures. For example, consider the case of the truck engine mentioned above, with the target scenario 8 between the maintenance to be performed 6 and the next scheduled maintenance 7, defined as follows: - Controllable planned operating conditions 110: average load transported of 2T, distance traveled of 100,000km, average speed of 90km / h, - Uncontrollable planned conditions of use 110, ambient air temperature 20°C, average gradient of the route used 5%.

[0227] In this example it is further assumed that the maximum limit 18 is for an average load of 2.5T, a distance travelled of 110,000km and an average speed of 100km / h.

[0228] In this example, as constructed, bounding polygon 181 depicts the following usage scenario 8 as meeting zero failures until the next scheduled maintenance 7: - The 2.5T was transported over an average of 100,000 kilometers between two maintenance operations, traveling at an average speed of 90 km / h, - Between two maintenance operations, the 2T traveled an average of 110,000 kilometers, traveling at an average speed of 90 km / h, - Between two maintenance operations, the 2T will travel at an average of 100km / h and will be transported over an average of 100,000km.

[0229] Nevertheless, as constructed, the limiting polygon 181 never makes it possible to tell whether a 2.3T, transported on average over 105,000 km between two maintenance operations and travelling on average at 95 km / h, is a usage scenario that complies with the intended zero failures until the end of the next scheduled maintenance 7. Thus, according to an embodiment, the invention foresees determining differently all usage scenarios that are compatible with zero failures until the date of the next maintenance. To do this, the system 1 includes a graphical representation in the form of a chart with at least one curve showing the maximum limit 18 of a first scheduled condition 110 as a function of at least a second one of said scheduled conditions 110. This chart makes it possible to determine a usage scenario that is compatible with zero failures for a scheduled period 70 comprised between the maintenance 6 to be performed and the next scheduled maintenance 7.

[0230] In other words, the invention envisages constructing at least one graphical representation of the nomogram type, as an example in the form of a set of curves on a chart (hereinafter "chart"), allowing richer information regarding the limit scenarios. For a given scenario 8 (defined by target values ​​of the controllable and non-controllable scheduled usage conditions 110), the constructed chart graphically represents a set of limit usage amounts that are compatible with zero failures until the date of the next scheduled maintenance 7, i.e. all n-tuples of the values ​​of the controllable scheduled usage conditions 110, each n-tuple indicating, for (n-1) controllable scheduled usage conditions 110 values, the maximum limit 18 of the n-th controllable scheduled usage condition 110 that is compatible with zero failures until the date of the next scheduled maintenance 7 (the non-controllable scheduled usage conditions 110 of the scenario 8 are locked to their respective values). For example, in the case of the truck engine mentioned above, as scenario 8 between the maintenance to be performed 6 and the next scheduled maintenance 7, a scenario defined by the following scheduled usage conditions 110 is considered: - Controllable planned operating conditions 110: average load transported of 2T, distance traveled of 100,000km, average speed of 90km / h, - Uncontrollable planned conditions of use 110, ambient air temperature 20°C, average gradient of the route used 5%.

[0231] In this case, with a graphical representation of the distance traveled as a function of the average load transported, the chart is a set of curves each associated with a given average speed between two successive maintenance operations. Each iso-speed curve is a set of triplets of values ​​of three controllable planned conditions of use 110 (average load transported). Each triplet defines a limit usage that meets zero failures until the date of planned maintenance 7 (locked to the respective values ​​of the uncontrollable planned conditions of use 110 (ambient air temperature and average gradient of the route used) of scenario 8, 20° C. and 5%). Thus, the curve associated with an average speed of 95 km / h defines, for a given value of the average load transported (for example, 2.3 T), the value of the maximum limit 18 of the distance covered until the next planned maintenance 7 (101,000 km in this example), i.e. the maximum possible value that meets zero failures until the date of planned maintenance 7 (locked to the respective values ​​of the uncontrollable planned conditions of use 110 (ambient air temperature and average gradient of the route used) of target scenario 8, 20° C. and 5%). To do this, in order to construct the curves associated with a given value of the controllable scheduled operating conditions 110, the present invention calculates, by iteration and by simulation enabled by the model 16, the maximum limit 18 of the controllable scheduled operating conditions 110 observed between the performed maintenance 6 and the scheduled maintenance 7 for the remaining different values ​​of the controllable scheduled operating conditions 110 (the non-controllable scheduled operating conditions 110 are still locked to their respective values ​​of the target scenario 8). In the truck engine example mentioned above, there are three controllable scheduled operating conditions 110. The chart represents a set of curves, each associated with a value of a first controllable scheduled operating condition 110 (e.g., average speed).

[0232] A maximum limit 18 of a second controllable scheduled condition 110 (e.g., the distance traveled) is calculated by simulation with the model 16 for different values ​​of a third scheduled condition 110 (in this example, the average load transported) in order to construct a curve related to the value of said first controllable scheduled condition 110 (e.g., the average speed of 100 km / h), while the non-controllable scheduled use conditions 110 (ambient air temperature and the average gradient of the route used) are always locked to the respective values ​​of the target scenario 8 (i.e., in this example, 20° C. and 5%, respectively). Thus, for example, as one possible limit usage, the chart is a triplet (2.3T; 101,000 km; 95 km / h). In FIG. 11, the aforementioned example of a machine 2 such as a truck engine is considered, with the use conditions 110 being the load P transported, the distance traveled d, the average speed V, the ambient air temperature T°, and the average gradient A of the route used. In FIG. 11, the intended state 130 is defined by the material condition indicator 120 corresponding to the average compression pressure U of the cylinder, and the value U of the minimum state 17 is set as the minimum required for the correct operation of the equipment 2. min 11 shows an example of such a chart representation of usage for equipment 2 associated with its fixation logs 9 and 11 for a given maintenance decision regarding a maintenance 6 to be performed, for a usage scenario 8 over a planned period 70 defined by target average values ​​(Pc, dc, Vc, Tc, Ac) for each of the controllable (P, d, V) and uncontrollable (T°, A) usage conditions 110 (recall from the previous example that Pc=2T; dc=100,000 km; Vc=90 km / h; T°c=20° C.; Ac=5%).

[0233] In Fig. 11, the chart shown is a set of three curves in a graphic representation of the distance traveled (d) as a function of the average load transported (P). The visible curves are associated with three average speeds (90, 95, 100 km / h) given during two successive maintenance operations. For two intended target average values ​​of the controllable operating conditions 110 P and V (Pc = 2.3T; Vc = 95 km / h), the graph shows the value of the maximum limit 18 (dlim = 101,000 km) for the last controllable operating condition 110 (d) as calculated by the model 16. Thus, the usage scenario defined by three intended target average values ​​for the controllable planned usage conditions 110 (Pc=2.3T, dlim=101,000km, and Vc=95km / h) and two target values ​​for the uncontrollable planned usage conditions 110 (Tc=20°C, and Ac=5%) corresponds to a limiting scenario that meets the intended zero failures until the end of the planned period 70, i.e., meets the planned state 130 of the equipment 2 at the end of the planned period 70 corresponding to the state 17 (U=Umin) minimally required for correct operation.

[0234] According to an embodiment, when the maintenance 6 is performed and in the middle of the scheduled period 70, the invention refreshes the scheduled state 130 of the equipment 2 at any time, such as the end of the scheduled period 70 or the day before the next scheduled maintenance 7. In other words, at a given time in the middle of the scheduled period 70 during which the scenario 8 of the equipment 2 is being performed, taking into account the usage that has actually been made to the equipment 2 since the beginning of the scheduled period 70, it is determined whether the equipment 2 at the time being considered, in the remaining situation of the scenario 8, retains a sufficient probability of zero failures during operation in the remaining middle until the end of the scheduled period 70 or until the day before the next scheduled maintenance 7 following the maintenance 6 to be performed. Furthermore, the invention already foresees the selection of the optimal maintenance decision from among the sufficient maintenance decisions on the day before the maintenance 6 to be performed, which allows a scheduled state 130 at the end of the scheduled period 70 compatible with the operation of the equipment 2 (taking into account the scenario 8). However, the present invention foresees refreshing a scheduled state 130 that is at the end of a scheduled period 70 at any point during the scheduled period 70, as this refresh may potentially be necessary.

[0235] In reality, the actual usage of equipment 2 may have differed from the predicted usage of scenario 8 during the elapsed portion of the planned operating period 70 between its initiation and the time considered.

[0236] Moreover, during the elapsed part of the scheduled period 70, the equipment 2 may be subject to abnormal operating transients, in particular penetrating into a near-destructive region or penetrating into an intermediate region between said near-destructive region and the region of normal operation. Moreover, the invention selects the ageing and therefore future behavior of the equipment affected by the collection of these abnormal transients. Thus, the invention foresees refreshing, during operation, the scheduled state 130, as it is at the end of the scheduled period 70, with the total number of abnormal transients that may occur from the beginning of the scheduled operating period 70 to that time. Similarly, if the invention foresees determining, on the day before the maintenance 6 is to be performed, the optimal determination of the maintenance 6 to be performed and the associated limit usage (i.e. the maximum possible usage observed during operation for zero failures after the maintenance 6 is performed until the date of the next scheduled maintenance 7), the invention foresees refreshing this limit usage during operation as potentially necessary.

[0237] In fact, the limiting usage of equipment 2 is likely to contract during operation due to the actual usage of equipment 2 during the elapsed portion of the planned period 70 (which is more limiting for equipment 2 than the originally predicted scenario 8) and due to any abnormal operating transients.

[0238] By refreshing during operation the limit usage adapted to zero failures until the next scheduled maintenance 7, the invention foresees allowing the user to identify the maximum limit 18 to which the usage of the equipment can be pushed without compromising zero failures, or to which the initially assumed usage must be limited in order to benefit from zero failures. For example, considering the case of the truck engine mentioned above, it turns out that the decision D6 of the maintenance 6 to be performed (i.e., replacement of the diesel filter and the injection pump) is taken and at the end of half the scheduled period 70, the average load transported correctly complies with the target value of 2T, the distance traveled (50,000 km) correctly complies with a proportional amount of the target value of 100,000 km at the end of the scheduled period 70, but the average speed is 105 km / h, exceeding the target value of 90 km / h of scenario 8 and the value of the maximum limit 18 of 100 km / h.

[0239] Under these conditions (and in addition to any abnormal transients in the course of the elapsed part of the planned period 70), continuing the operation of the truck in the context of the initial scenario 8 with its target values ​​(2T, 100,000 km, 90 km / h) does not meet zero failures until the date of the next target planned maintenance 7. It is therefore desirable to refresh the calculation of the maximum limits 18 of the various planned operating conditions 110 that the operator must observe in order to benefit from zero failures until the end of the planned period 70. To do this, after the maintenance 6 in question has been performed, updated values ​​are submitted to the model 16 in question. These values ​​correspond, as before, to at least one of the tasks 90 of the production and maintenance log 9 and to at least one task 90 of the maintenance 6 to be performed.

[0240] These values ​​also correspond to at least one of the usage conditions 110 in the usage log 11 after the maintenance 6 is performed. These values ​​also correspond to the planned usage conditions 110 of the scenario 8.

[0241] Based on these values, the model 16 then refreshes the intended state 130 of the device 2. Furthermore, the intended state 130 is compared to the minimum state 17 identified as necessary for the operation of the device 2. In summary, according to one embodiment, after the maintenance 6 has been performed, following the maintenance 6: the value of at least one of the tasks 90 of the production and maintenance log 9 and the value of at least one of the tasks 90 of the maintenance to be performed 6, At least one value of the usage conditions 110 of the usage log 11 after the maintenance 6 is implemented; - the values ​​of the intended conditions of use 11 of the scenario 8 are submitted to the model 16, The model 16 refreshes the planned state 130 of the equipment 2, which is compared to the minimum state 17 identified as necessary for the operation of the equipment 2. In other words, at a given time during the next scheduled period 70, a prediction of the planned state 130 of the equipment 2 is refreshed for what will be the end of the scheduled period 70 or the day before the scheduled maintenance 7. To do this, and for the time considered, when the maintenance 6 has been performed and the equipment 2 is operational, the model 16, as a customized simulator for the equipment 2, considers the arguments submitted as input to the model 16, namely: - Device 2 Manufacturing and Maintenance Log 9, - Task 90 carried out with regard to Conservation 6, - the usage log 11 up to the time under consideration, followed by the usage during the elapsed part of the planned operating period 70 up to the time under consideration, - by adapting the usage of the equipment 2 foreseen in the remainder of the scenario 8, from that point on, for the remainder of the planned operating period 70, until the day before the next planned maintenance 7. Thus, this possibility of being able to refresh the planned state 130 of the equipment 2 so that it is at the end of the planned operating period 70 after a future execution of the entire scenario 8 (taking into account the tasks 90 actually performed for the maintenance 6 and taking into account the actual usage of the facility 3 and the equipment 2 from the start of the planned operating period 70 until the considered point in time) allows the invention to make various applications of the model 16 such that the maintenance 6 is foreseen the day before it is to be performed.

[0242] In particular, the usage of the model 16 makes it possible to determine, at any time during the planned period 70, whether the continuation of the scenario 8 remains compatible with zero failures until the end of the planned period 70, by comparing the planned state 130 with the minimum state 17 required for the operation of the equipment 2.

[0243] In particular, the usage of the model 16 allows for refreshing at any time during the forecast period 70 the maximum limit 18 of the limited usage that fits zero failures until the next scheduled maintenance 7, the limit polygon 181, the usage margin before failure (MUBF), as well as a chart of all limit scenarios, which allows the user to identify limits to which the usage of the equipment 2 can be pushed without compromising zero failures, or limits to which the initially assumed usage must be limited in order to benefit from zero failures. At a given time during the forecast period 70, the invention refreshes the planned state 130 of the equipment 2, such as at the end of the forecast period 70 and after the execution of scenario 8. Similarly, at a given time during the forecast period 70, the invention refreshes the planned state 130 of the equipment 2 for the period between that time and the end of the forecast period 70, such as at the end of the forecast period 70 and after the execution of a scenario different from the initially predicted usage scenario 8.

[0244] Instead of submitting the usage of the equipment 2 in the context of scenario 8 among the arguments submitted as input to the model 16, the usage of the equipment 2 in the context of a scenario not foreseen is actually submitted to the model 16 from the time point, for the remainder of the planned operating period 70, until the day before the planned maintenance 7. Similarly, the invention refreshes the maximum limit 18 of the limited usage that fits zero failures until the next maintenance 7, the limit polygon 181, the usage margin before failure (MUBF), as well as the chart of all limit scenarios, between the time point and the end of the planned period 70, with the aim of executing a scenario different from the initially foreseen usage scenario 8, so that the user can identify the limits to which the usage of the equipment can be pushed without compromising zero failures, or the limits to which the initially assumed usage must be limited in order to benefit from zero failures, with the aim of executing a scenario different from the initially foreseen usage scenario 8. Thus, the present invention enables the operator to determine, during a planned period 70, the amount of usage that the operator can or should perform on the equipment 2 in order to maintain this amount of usage that complies with the zero failure requirement until the next maintenance date, both in the context of the originally predicted scenario 8, as well as in the context of an unforeseen scenario that replaces the originally predicted scenario 8.

[0245] Advantageously, the simulation by the model 16 further allows for a stepwise determination of sufficient maintenance decisions for successive planned maintenance operations 7 in the life of the equipment 2, taking into account the target usage (i.e. successive scenarios 8) intended by the operator over the short term as well as over the long term. Thus, the system 1 allows for constantly determining different possible sufficient maintenance schedules throughout the entire useful life of the equipment 2 for the intended usage profile. To do this, according to one embodiment, at least the following steps are implemented:

[0246] Firstly, it is assumed that at least said sufficient determination of the maintenance 6 to be performed has been performed, and scenario 8 is assumed to have been performed up to the scheduled maintenance 7 which follows said maintenance 6 to be performed.

[0247] Then, at least one variation is made to at least one of the values ​​of the tasks 90 of the scheduled maintenance 7. When the values ​​are submitted to the model 16, the values ​​of the variations are introduced as well as the values ​​of the next scenario 81 that is predicted for the next scheduled period 710 (i.e. the operating period between the scheduled maintenance 7 and the next scheduled maintenance 71).

[0248] In all variants, at least one sufficient decision is selected at the time of a scheduled maintenance 71 following the scheduled maintenance 7 if the scheduled state 130 of the equipment 2 is equal to or greater than the minimum state 17 for that scheduled maintenance 7 .

[0249] Once this sufficiency decision has been selected, the steps are repeated following every other scheduled maintenance in the life of the equipment. In short, once a sufficiency decision has been defined for a maintenance to be performed 6, a series of sufficiency decisions are determined recursively for the scheduled maintenance operations 70 following the maintenance to be performed 6. Indeed, on the day before the maintenance to be performed 6 (taking into account the maintenance decisions determined to be sufficient for the maintenance to be performed 6 and assumed to have been performed, and taking into account the scenario 8 assumed to have been performed), the model 16 makes it possible to determine a sufficiency decision for the next scheduled maintenance 7 following the maintenance to be performed 6. In this sense, the model 16 makes it possible to iterate in order to determine the sufficiency decisions step by step from one scheduled maintenance operation 7 to the next scheduled maintenance operation 7.

[0250] For this reason, as before, for each possible maintenance decision for a scheduled maintenance 7 following a performed target maintenance 6, the scheduled state 130 of the equipment 2 is simulated for what it will be after the next scenario 81, i.e. at the end of the next scheduled period 710, the day before the next scheduled maintenance 71 (i.e. following the target scheduled maintenance 7). To do this, taking into account these assumptions, the model 16 calculates the model 16 based on the arguments submitted as input to the model 16, namely: - the manufacturing and maintenance log 9 of the device 2 (up to the day before the maintenance 6 to be performed), as amended with the task 90 assumed to have been performed with respect to the maintenance 6 to be performed; - task 90 of possible maintenance decisions of the object with respect to planned maintenance 7, - Usage log 11 (up to the day before the maintenance of the object 6 is performed) followed by scenario 8 assumed to have been performed during the planned period 70; - used as a simulator customized for the equipment 2 by adapting the next scenario 81 of the equipment 2 (in the middle of the next scheduled period 710 and up to the day before the next scheduled maintenance 71). If the scheduled state 130 of the equipment 2 thus simulated by the model 16 for the time corresponding to the end of the next scheduled period 710 (i.e. the day before the next scheduled maintenance 71) complies with the operation criteria of the equipment 2, compared with the minimum state 17 required for the operation of the equipment 2, then the maintenance decision regarding the scheduled maintenance 7 following the maintenance 6 to be performed can be considered sufficient to give the equipment 2 a sufficient probability of zero failures during operation in the context of the next scenario 81, in the middle and at the end of the next scheduled period 710. In this way the recursion is established. Indeed, on the day before the to-be-performed maintenance 6 (maintenance of rank j) (considering the maintenance decision determined to be sufficient and assumed to have been performed for the to-be-performed maintenance 6, and considering the scenario 8 assumed to have been performed during the next scheduled operating period 70), the model 16 made it possible to determine a sufficient maintenance decision for the scheduled maintenance 7 (maintenance of rank j+1), i.e. a decision that authorizes zero failures for the next scenario 81 during and by the end of the next scheduled period 710. Furthermore, the invention makes it possible to determine a sufficient decision for the to-be-performed maintenance 6. Thus, the simulation by the model 16 makes it possible to determine step by step, by recursion, the sufficient maintenance decision sequence for any scheduled maintenance 7, up to the intended rank of the scheduled maintenance 7 (maintenance of rank j+n) in the life of the equipment 2. In Figs. 12 and 13, the equipment 2 associated with the logs 9 and 11 of the equipment 2 is taken into account, with the usage scenario 8 (in Fig. 12) during the scheduled period 70, followed by the next scenario 81 (in Fig. 13) during the next scheduled period 710. 12 and 13 show a step-by-step simulation of the planned state 130 of the equipment 2 on the day before each scheduled maintenance operation 7, 71 and the iterative determination of a sufficient maintenance decision for said scheduled maintenance operation 7, 71. The model 16 makes it possible to determine a sufficient maintenance decision for a maintenance operation 6 to be performed (of rank j), which then makes it possible to determine a sufficient maintenance decision for the next scheduled maintenance 7 of rank j+1.First, FIG. 12 shows that for a possible maintenance decision for the maintenance 6 to be performed, the model 16 makes it possible to predict the planned state 130 of the equipment 2 at the end of the planning period 70. In particular, the model 16 therefore makes it possible to determine whether said possible decision is sufficient to allow the intended zero failures until the end of the planning period 70 (depending on whether the planned state 130 of the equipment 2 at the end of the planning period 70 is sufficient for the minimum state 17). The model 16 obviously makes it possible to determine a sufficient maintenance decision among the possible maintenance decisions for the maintenance 6 to be performed. In a second step, FIG. 13 shows that for a maintenance decision identified as sufficient for a maintenance work of rank j (in this case for the maintenance 6 assumed to have been performed) and for a possible maintenance decision for the next maintenance of rank j+1 (for the next scheduled maintenance 7, if at hand). Thus, model 16 allows to determine whether this latter possible decision is sufficient to enable the intended zero failures until the end of the next scheduled period 710 (according to whether the scheduled state 130 of equipment 2 at the end of the next scheduled period 710 is sufficient for the minimum state 17). Thus, Figure 13 shows that for a maintenance decision identified as sufficient for a maintenance action of rank j, model 16 allows to determine a sufficient maintenance decision among the possible maintenance decisions for a subsequent maintenance action of rank j+1.

[0251] As can be seen in Figures 12 and 13, the system 1 thus allows to determine by recursion the sufficient maintenance decisions step by step from one maintenance action to another in the life of the equipment 2, i.e. from a maintenance action of rank k to a maintenance action of rank k+1. In particular, Figures 12 and 13 show how the model 16 is used both for determining the planned state 130 of the equipment 2 in the course of a scheduled period 70 and for determining the planned state 130 of the equipment 2 in the course of a next scheduled period 710. Figures 12 and 13 in fact show the configuration of the input arguments of the model 16 in both cases. In particular, due to the fact that the step-by-step simulation of the planned state 130 on the day before each scheduled maintenance 7, the simulation of the planned state 130 of the next scheduled period 710 after the maintenance of rank j+1 (as can be seen in Figure 13) integrates the maintenance decisions that allowed the simulation to identify as sufficient for the maintenance of rank j (as can be seen in Figure 12). Similarly, the invention determines step by step the successive planned states 130 of the equipment 2 during two planned periods 70, 710. For this reason, two maintenance decisions are taken into account, one for the performed maintenance 6 (maintenance of rank j) and one for the planned maintenance 7 (maintenance of rank j+1), and two usage scenarios 8, 81 for each of said planned periods 70, 710. In this regard, Fig. 14 shows two example curves of the planned states 130 of an equipment 2 such as a pump (reduced in this example to its "flow rate Q" material indicator 120).

[0252] The model 16 makes it possible to predict the planned state 130 of the pump for various points in time of the planned period 70 (from the maintenance of rank j to the maintenance of rank j+1) and then for various points in time of the subsequent planned period 710 (from the maintenance of rank j+1 to the maintenance of rank j+2). In particular, the simulation of the model 16 highlights the need to work on the pump during the maintenance of rank j+1 in order to benefit from the intended zero failures (otherwise the pump would fail before the maintenance of rank j+2, as can be seen by the dotted extension of the curve).

[0253] FIG. 14 particularly highlights the discontinuity of flow on either side of the maintenance of rank j+1, due to a maintenance decision made on the pump (on the day of said maintenance of rank j+1). This maintenance decision of rank j+1 is moreover sufficient, since it allows sufficient flow until the day before the maintenance of rank j+2. Similarly, and more generally, the invention determines stepwise the successive planned states 130 of the equipment 2 in the course of as many successive planned periods 70, 710 as are intended to be considered in the life of the equipment 2. For this reason, a series of maintenance decisions is considered, respectively, for the maintenance 6 to be performed (maintenance of rank j) and then for the various planned maintenance operations 7, 71 (maintenances of rank j+1 to j+k) preceding the planned period 70, 710 of interest, and a series of usage scenarios 8, 81 is considered for each of the planned periods 70, 710 of interest. In this regard, FIG. 15 shows an example of two overall curves of a planned state 130 of an equipment 2, such as a pump (reduced in this example to its "Flow Q" material indicator 120).

[0254] The model 16 makes it possible to predict the planned states 130 of the pump for various instants of time in the period between the maintenance of rank j and the maintenance of rank j+k, and for various instants of time in the period between the maintenance of rank j+k and the maintenance of rank j+n. These curves are obtained from the prediction, by simulation of the model 16, of the successive planned states 130 of the pump for each operating period between two successive maintenance operations (the intervals delimited by the vertical dotted lines in FIG. 15).

[0255] It is therefore emphasized that work needs to be done on the pump during the maintenance of rank j+k in order to benefit from the intended zero failures during each operation period between the maintenance of rank j and the maintenance of rank j+n.Similarly, Fig. 15 particularly highlights the discontinuity of flow rates on either side of the maintenance of rank j+k due to the maintenance decision to be made on the pump (on the day of said maintenance of rank j+k).

[0256] Thus, the present invention enables determining the planned state 130 of the equipment 2 to be at any point in the course of any subsequent planned period 710 in the life of the equipment 2 (in the context of a given subsequent scenario 81 and taking into account a maintenance decision that has been determined to be sufficient for scheduled maintenance 7).

[0257] Thanks to this predictive capability, as before, the present invention makes it possible to identify, for each subsequent planned period 710 in the life of the equipment 2, limits to which the equipment usage can be pushed without compromising the intended zero failures, or limits to which the initially envisaged usage should be limited in order to benefit from zero failures.

[0258] Indeed, for each next scheduled period 710 in the life of the equipment 2 (taking into account the maintenance decisions determined to be sufficient for the corresponding service maintenance 7 after a given next scenario 81), the invention determines: - determining the maximum limits 18 of the controllable intended conditions of use 110 as described above; - therefore, limit the usage limit that conforms to zero failures in the next scenario 81 during the course of the next scheduled period 710, and do so until the day before the next scheduled maintenance 71; - these maximum limits 18 of the planned operating conditions 110 thus calculated are then used to quantify, as before, the suitability of the sufficient maintenance decision (for the task 90 for the planned maintenance 7) for the next scenario 81, preferentially by a Margin of Use Before Failure (MUBF) or by a deviation 182 between a target value of the next scenario 81 and a maximum limit 18 of at least one controllable planned operating condition 110 of the next scenario 81; - As before, a chart is constructed showing the set of limit scenarios valid for the next scheduled period 710. The simulation by the model 16 thus makes it possible to determine, recursively, stepwise, a series of sufficiency decisions for any scheduled maintenance 7, up to the rank of the intended scheduled maintenance 7. In addition, for each subsequent scheduled period 710, the simulation made possible by the model 16 makes it possible to determine, for each of the sufficiency decisions, the maximum limit 18, the limit polygon 181, the margin of use before failure (MUBF), the chart of limit scenarios for each controllable scheduled operating condition 110 of the corresponding scenario 81.

[0259] According to an embodiment, the optimal decision is selected from at least the sufficient decisions for the corresponding maintenance. In other words, for each planned maintenance step 7 in the life of the equipment 2, an optimal maintenance decision is selected from among the maintenance decisions determined to be sufficient, as described above, and maintenance 6 is performed. The optimal decision is preferentially selected with respect to a usage margin before failure (MUBF) allowed by each sufficient maintenance decision considered, or with respect to at least one deviation 182 between a target value of a scenario 8 and a value of a maximum limit 18 for at least one of the planned operating conditions 110 of said scenario 8.

[0260] Thus, the present invention foresees optimizing each of the scheduled maintenance operations 7 in the life of the equipment 2 by determining an optimal maintenance decision for each maintenance step in the life of the equipment 2, taking into account successive scenarios 8 in the life of the equipment. This decision on the nature of the task 90 for the scheduled maintenance 7 is customized for the complete log of the equipment and suited to the scenario 8 to allow zero failures until the end of the scheduled period 70, allowing the margin of use before failure (MUBF) intended by the operator. In other words, the present invention determines by recursion a series of optimal maintenance decisions for each of the scheduled maintenance operations 7 in the remaining life of the equipment 2. Furthermore, for each scheduled period 70 in the life of the equipment 2, the present invention determines the maximum limit 18 (for each controllable scheduled usage condition 110 of the corresponding scenario 8), the limit polygon 181, the margin of use before failure (MUBF), the chart of the limit scenarios, taking into account the maintenance decision identified as optimal. Thus, for each planned period 70 in the life of the equipment 2, the invention makes it possible to identify limits to which the equipment usage can be pushed without compromising zero failures, or limits to which the initially envisaged usage should be limited in order to benefit from zero failures. In other words, the invention recursively determines the limit usage of the series of equipment 2 for each planned period 70 in the remaining period of the equipment 2's life.

[0261] As mentioned above, for the maintenance 6 to be performed, the model 16 makes it possible to distinguish between sufficient and insufficient maintenance decisions. In the case of a sufficient maintenance decision, the maintenance decision is suitable for the usage scenario 8 in that it is sufficient to tolerate zero failures until the end of the planned period 70. In the case of a maintenance decision determined to be insufficient, the operation of the equipment 2 in the context of the usage scenario 8 leads to a failure of the equipment 2 before the end of the planned period 70. In the latter case, the calculation of a failure date 19 is justified. Thus, according to an embodiment, if the maintenance decision is insufficient and the planned state 130 (during the planned period 70) is less than the minimum state 17 in question, a failure date 19 is determined for the insufficient maintenance decision.

[0262] The failure date 19 corresponds to the time when the planned state 130 is equal to the minimum state 17. A simulation by the model 16 makes it possible to determine this failure date 19. To do this, for a given decision of insufficient maintenance with respect to the maintenance 6 to be performed, for a planned period 70 and its scenarios 8, the following procedure is carried out: a given point in time is considered in the course of the planned period 70 and the sub-scenario 80 of the scenario 8 and corresponds to that point in time. The model 16 then performs a simulation based on the arguments submitted as input to the model 16, namely: - Device 2 Manufacturing and Maintenance Log 9, - Insufficient conservation judgments taken into account with respect to the conservation of the object to be implemented6; - Equipment 2 Usage Log 11, - used as a simulator customized for the equipment 2 by adapting the usage of the equipment 2, which is foreseen in the context of the defined partial scenarios 80 up to the considered time point of the planned period 70. The model 16 is then used to solve, for that time point, the equation in which the planned state 130 of the equipment 2 at that time point corresponds to the minimum state 17. For the considered insufficient maintenance decision and for the scenario 8, the failure date 19 is determined to be the solution of that equation.

[0263] The failure date 19 of the equipment 2 is thus determined due to the assumed insufficient maintenance decision for the to-be-performed maintenance 6. The same procedure can be applied to determine the failure date 19 related to the next scenario 81, related to the insufficient maintenance decision for the scheduled maintenance 7, during the next scheduled period 710. FIG. 16 considers an equipment 2, such as a pump, associated with its logs 9 and 11, a scenario 8 in the course of a scheduled period 70 (included between the to-be-performed maintenance 6 (of rank j) and the next scheduled maintenance 7 (of rank j+1)), a partial scenario 80 of the scenario 8 associated with a time point, as well as the insufficient maintenance decision considered for the to-be-performed maintenance 6. FIG. 16 shows the curves of the scheduled state 130 of the pump (reduced in this example to its "flow rate Q" material indicator 120), which allows the model 16 to predict for various times of the scheduled period 70.

[0264] The curve of the planned condition 130 over time decreases due to aging until it reaches a planned condition 130 corresponding to the minimum condition 17 prior to the non-date of the scheduled maintenance 7. In the illustrated case, the possible maintenance decision is in fact an insufficient decision and does not allow a planned condition 130 greater than the minimum condition 17 until the end of the scheduled period 70. The intercept of the curve with the minimum condition 17 corresponds to the failure date 19 of the pump.

[0265] According to one embodiment, for a maintenance 6 to be performed without a maintenance decision identified as sufficient, and for at least one given insufficient decision of the maintenance 6 to be performed, the end date of the useful life of the equipment 2 is determined to be the failure date 19 associated with the maintenance decision. In other words, if all possible maintenance decisions are found to be insufficient for the maintenance 6 to be performed (in the context of scenario 8, to enable zero failures during and by the end of the planned period 70), the equipment 2 is at the end of its life and the maintenance 6 to be performed corresponds to the last maintenance in the life of the equipment 2. When the maintenance 6 to be performed is thus identified for a given maintenance decision as the last maintenance in the life of the equipment 2, the end date of the useful life of the equipment 2 is determined to be the failure date 19 associated with the maintenance decision.

[0266] Once a scheduled maintenance job 7 is identified as the final maintenance in the life of the equipment 2, the same procedure is applied to determine the end date of the equipment 2's useful life associated with the given maintenance decision and the next scenario 81.

[0267] According to one embodiment, for a target maintenance 6 (or planned maintenance 7) identified as the final maintenance in the life of the equipment 2, a maintenance decision that optimizes any combination of the failure date 19, margin of end use (MLU), and task 90 constraints of the final maintenance decision is selected from among the possible maintenance decisions.

[0268] In other words, all possible maintenance decisions are considered for the maintenance 6 to be performed, which is identified as the final maintenance in the life of the device 2. For each of these maintenance decisions, the minimum operating period (D MIN ), i.e. a minimum operating period determined as a function of the constraints of the tasks 90 for the maintenance decision (i.e. as a function of the nature and amount of the tasks 90).

[0269] For example, consider the case of equipment 2, such as a truck engine. Five possible maintenance decisions are identified for the final maintenance in the life of equipment 2, and for each, the intended minimum operating period (D MIN ) is determined, and the five possible conservation decisions are, namely: - Change the diesel filter (during the intended minimum operating period D equal to 3 days). MIN ) Judgment E1, - Decision E2 to replace injectors and diesel filters (intended minimum operating period D MIN is equal to 1.1 months), - (intended minimum operating period D equal to 2 months MIN 4) Decision E3 to replace the injection pump and injector; - Decision E4 to replace the injection pump, injectors and diesel filters (intended minimum operating period D MIN is equal to 2.1 months), - Injection pumps, injectors, diesel filters and cylinder segmentation (intended minimum operating period D equal to 12 months) MIN Among the possible maintenance decisions, each decision is then determined so that the service life extension (SLE) associated with the maintenance decision (i.e., the period between the last maintenance in the life of the equipment 2 and the end date of the service life of the equipment 2) is equal to or exceeds the minimum operating period (D MIN ) (i.e., SLE>D MIN ).

[0270] For example, the case of the truck engine mentioned above is considered with five possible maintenance decisions identified for the last maintenance to be performed during the engine's life, and for each of them, the minimum intended operating period (D MIN ) and the Service Life Extension (SLE) are determined, and five possible conservation decisions are given, namely: - Decision E1 to change the diesel filter (intended minimum operating period of 3 days D MIN and 1-month extended SLE, in which the injector is the limiting factor). - Decision E2 to replace injectors and diesel filters (intended minimum operating period of 1.1 months D MIN and 1.5 months of extended SLE, with the injection pump being the limiting factor). - Decision E3 to replace the injection pump and injectors (intended minimum operating period of 2 months D MIN and 0.5 months life extension SLE, where the diesel filter is the limiting factor). - Decision E4 to replace the injection pump, injectors and diesel filters (2.1 months intended minimum operating period D MIN and 6-month extended SLE, with segmentation being the limiting factor). - Judgment E5 (intended minimum operating period of 12 months) for replacing injection pumps, injectors, diesel filters, and cylinder segmentation MIN and 7 months of extended SLE, with the crankshaft being the limiting factor).

[0271] Judgment E3 is the minimum intended operating period D MIN The same is true for decision E5. The decisions selected are therefore decisions E1, E2 and E4, which, conversely, do not allow the service life extension period SLE (0.5 months) to conform to the intended minimum operating period D (2 months). MIN For each possible maintenance decision thus selected for the final maintenance in the life of the equipment 2, the present invention determines for each controllable planned use condition 110 of the scenario 8: - the intended minimum operating period D of the device 2 after the maintenance to be carried out; MIN The invention calculates this final usage limit in a similar manner to the maximum limit 18. The final usage limit is therefore determined by the intended minimum operating period D of the device 2. MIN The limit usage observed over the course of the forecast period 70 for zero failures by the end of the forecast period. - the deviation 182 between the target value and the final usage limit value in scenario 8 for that controllable planned usage condition 110;

[0272] Therefore, for each possible maintenance decision selected, the Margin of Ultimate Usage (MLU) is estimated for all controllable planned usage conditions 110 of scenario 8 as being the average of the deviations (i.e., in a similar manner to the calculation of Margin of Usage Before Failure (MUBF)).

[0273] For example, the case of the truck engine mentioned above is considered, and three possible maintenance decisions (E1, E2, and E4) are selected for the final maintenance in the life of the equipment, and for each of them, a Margin of End Use (MLU) is calculated, and the maintenance decisions are, i.e., - Decision E1 to change the diesel filter (with a margin MLU of 3%), - Decision E2 to replace injectors and diesel filters (with 4% margin MLU), - Decision E4 to replace the injection pump, injectors and diesel filters (with a margin MLU of 12%). Then, from among the previously selected last maintenance decisions, the best last maintenance decision is determined as the last maintenance decision, but without limitation, - Optimize Service Life Extension (SLE) - Optimize Margin of End Use (MLU) - or the applicable service life extension period SLE and the applicable minimum operating period D MIN The final maintenance efficiency (LME) is defined as the difference between the life extension period SLE and the maintenance efficiency (SLE) of the vehicle.MIN / SLE), or optimizing any other criteria and, at the operator's discretion, combining at least two of the criteria: useful life extension (SLE), end-use margin (MLU) and end-maintenance efficiency (LME); - Otherwise, if there is no maintenance decision that meets the aforementioned criteria, the decision is to replace the device 2 with a new one.

[0274] For example, the truck engine case mentioned above is considered, and its three possible maintenance decisions (E1, E2, E4) are MIN (to enable the last maintenance in the life of the equipment 2), and the maintenance decision is, i.e., - Decision E1 to change diesel filters (with 1 month extended SLE, 90% efficiency LME, and 3% margin MLU), - Decision E2 (with 1.5 months extended SLE, 26% efficiency LME, and 4% margin MLU) to change the injectors and diesel filters; - Decision E4 (with 6 month extended SLE, 65% efficiency LME and 12% margin MLU) to replace the injection pump, injectors and diesel filters.

[0275] Furthermore, in this example, it is chosen to select the optimal decision for the final maintenance according to criteria that simultaneously optimize service life extension (SLE), margin of end use (MLU) and final maintenance efficiency (LME).

[0276] Decisions E1 and E2 make no sense from an operational point of view (life extension SLE is only 1 and 1.5 months). Moreover, these decisions provide an excessively low reliability of use before failure (with margin MLU of 3% and 4%). Conversely, decision E4 makes much more sense operationally with a life extension SLE of 6 months, an acceptable reliability of use before failure (margin MLU>10%), and a very acceptable efficiency LME (65%).

[0277] Therefore, at the date of this last maintenance in the life of the engine, it is reasonable not to replace the engine as new, but to perform a last maintenance decision E4 (i.e. including replacement of the injection pump, injectors and diesel filters). The invention therefore foresees indicating the service life extension period of the engine. Thus, for a target maintenance 6 identified as the last maintenance in the life of the equipment 2, the invention selects, among the possible maintenance decisions, the service life extension period SLE (taking into account the failure date 19), the end-of-life margin (MLU) and the last maintenance efficiency LME (for duration D MIN , thus taking into account the constraints of said last maintenance task 90. ​​Thus, the present invention foresees prescribing a life extension period for the equipment 2, and thus determines an optimal life period for the equipment 2. Note that this prescribing for extending the equipment's life takes into account both the manufacturing and maintenance log 9 and the usage log 11 of the equipment 2, as well as the usage scenario 8 of the equipment 2 in the context of the life extension period of the equipment 2.

[0278] The same procedure can be applied to determine the optimal determination of the final maintenance for any scheduled maintenance 7 identified as the final maintenance in the life of the equipment 2 and associated with the next scenario 81.

[0279] According to one embodiment, the invention determines an optimal decision for manufacturing 4 the equipment 2, i.e. a critical task 90 of the manufacturing process 4 that optimizes the planned state 130 of the equipment 2 for a scenario 8 at the end of the planned period 70. For this reason, for at least two dummy equipment of the same series of the equipment 2, a separate manufacturing decision is associated, a simulation is performed by submitting said at least two production decisions and at least one planned use scenario 8 to said model 16; the model 16 generates at least one predetermined state 130 for each of the two dummy devices; - one of the at least two manufacturing decisions is selected in response to a predetermined state 130 of the two dummy devices; - The selected manufacturing decisions are accessible to the designer / manufacturer. In other words, the present invention considers various possible manufacturing decisions 4 (i.e. various possible combinations of critical tasks 90 for that manufacturing process 4). One dummy device in the series is associated with each combination.

[0280] For example, consider the case of equipment 2, such as a pump, where a critical task 90 in a manufacturing process 4 is reduced to the following tasks 90: - "Position and select pump bearings" with two possible options for the type of bearing: types A and B, - "Position and select pump impeller" with three possible options for impeller type: types A, B and C.

[0281] Thus, the various possible combinations of the critical tasks 90 of the manufacturing process 4 (i.e. the various possible decisions of the manufacturing process 4) are six combinations [type of bearing; type of impeller]: [A; A][A]; B][A]; C][B]; A][B]; B][B]; C]. Thus, various dummy machines are considered, each associated with a manufacturing decision 4 from among the various possible combinations of manufacturing 4, assumed to have been performed in the facility 3 and assumed to have operated during a first operating period 5 (reduced to a single operating period, i.e. not including any pre-emptive maintenance 10) in the context of a usage log 11 (identical for all dummy machines considered). In addition to a maintenance decision regarding a maintenance 6 to be performed (identical for all dummy machines considered), a scenario 8 (identical for all dummy machines considered) is considered.

[0282] For each dummy machine, the present invention uses the model 16 as a simulator customized for that machine to determine a planned state 130 of that machine at the end of a planned period 70 by fitting arguments submitted as input to the model 16, the arguments being: - Manufacturing and maintenance log 9 (reduced to a manufacturing decision of 4 specific to the dummy device), - a security judgment on the security 6 to be performed (common to all dummy devices considered); - Usage Log 11 (common for all dummy devices considered), - Scenario 8 (common for all dummy devices considered).

[0283] Thus, for all dummy machines associated with each production decision 4 (and any other parameters considered), the planned states 130 of the dummy machines are compared at the end of the planned period 70. When the planned states 130 are compared, the best planned state 130 represents the best production decision 4 (the best combination of critical production 4 tasks 90).

[0284] According to one embodiment, the simulation with the model 16 determines an optimal production decision 4, i.e. a task 90 of optimizing the optimal life cycle of the equipment 2 (the optimal life cycle includes a series of optimal maintenance decisions for the maintenance 6 to be performed for each scheduled maintenance operation 7, 71 in the life of the equipment 2, a series of limit usage (i.e. maximum limit 18) for each scheduled period 70, 710 in the life of the equipment 2, and an optimal useful life of the equipment 2), for a given usage profile (a set of a series of scenarios 8, 81 for different scheduled periods 70, 710 in the life of the equipment 2). To do this, for at least two dummy equipment of the same series of the equipment 2 associated with two distinct production decisions 4: - a recursive simulation is performed for each of the at least two manufacturing decisions 4 to determine an optimal maintenance decision for the series, a maximum limit 18 for the series, and an optimal service life for the dummy equipment; the optimal manufacturing decision 4 is selected as a function of the results of said simulation, - the selected optimal manufacturing decision 4 is accessible to the designer / manufacturer in question. In other words, the model 16 is able to simulate the planned states 130 of dummy equipment of the same series in the same way as for real equipment 2 (as seen before), so that the invention determines the optimal life cycle of said dummy equipment in the same way as for real equipment 2. Preferably, as many optimal life cycle simulations as there are dummy equipment (i.e. as many simulations as there are possible manufacturing decisions 4) are thus performed. The optimal manufacturing decision 4 is then selected depending on the results of said simulations, i.e. as the one that authorizes the best optimal life cycle. Thus, the system 1 makes it possible to determine a manufacturing decision 4 that optimizes the complete life cycle of the equipment 2 of the series, for the benefit of the manufacturer. According to another embodiment, the model 16 also makes it possible to evaluate the sensitivity of the behavior of the equipment 2 (in particular the planned states 130, the optimal life cycle) to the conditions of use 110 and to identify optimal values ​​for each condition of use 110, which shows a great number of possible approaches to increase the protection of the equipment against the conditions of use 110, or even to optimize the operating point of the equipment.

[0285] According to one embodiment, the invention offers operators a global monitoring of all the equipment 2 of a series they use, incorporated in the fleet (e.g. fleet of vehicles) of their facility 3. To do this, the system 1 is applied to a fleet of several equipment 2 of said series belonging to a single operator. The results obtained for each of said equipment 2 are combined, said results being then accessible at least to said operator.

[0286] In other words, the system 1 distributes to a given operator and constantly refreshes information regarding the optimal life cycle of each piece of equipment 2 of their fleet for one or more intended usage profiles (i.e. all of the usage scenarios 8, 81 over the short and long term of the useful life of each piece of equipment 2), i.e. in particular: - the optimal schedule of future maintenance for each piece of equipment 2 (i.e. the optimal maintenance decision of the series for the set of maintenance to be performed 6 and planned maintenance operations 7, 71 in the useful life of each piece of equipment 2), - the possible limit usage of the series for each device 2 over the long term of its useful life (i.e. in the useful life of each device 2 and for each planned period of operation 70, 710 in the context of scenarios 8, 81), - Refresh the optimal service life of each piece of equipment 2. Then, the system 1: - the nature of the parts required for each of the maintenance operations to be carried out in the life of each device 2 (i.e. for the maintenance operations to be carried out 6 and all future scheduled maintenance operations 7, 71 in the useful life of each device 2) for the benefit of the actors in the "supply chain"; - for the benefit of maintenance engineers, the scheduling of industrial maintenance (the nature of the tasks 90 to be performed for the maintenance 6 to be performed and all future scheduled maintenance work 7, 71 in the useful life of each piece of equipment 2); - for the benefit of the operator, all constraints (including the total cost) of future maintenance work in the life of each piece of equipment 2 (for the maintenance work 6 to be carried out and for all future scheduled maintenance work 7, 71 in the useful life of each piece of equipment 2), - for the benefit of the operator, it allows to timely predict and constantly refresh the scheduling of the possible future usage of each device 2, as well as the scheduling with respect to previously determined usage limits, over a long period of its useful life. Thus, the system 1 in particular: - constantly evaluating and refreshing the residual value of each device 2 during its lifespan; - determining an objective price for the asset that each piece of equipment 2 constitutes, and thus appropriately informing the decision to keep or resell the equipment 2. Consider the example of a fleet of equipment 2, such as trucks. The system 1 enables the operator of that fleet of trucks to answer the following questions: - What is the maximum possible usage of several new trucks by an operator between two successive maintenance operations for different operating profiles (e.g. in terms of the loads transported, the distances travelled, the average speed, the average gradient of the route used, the ambient air temperature) associated with different geographical areas (e.g. Russia, Senegal, France, etc.)? - An operator has several trucks that have reached halfway through their useful life and which they intend to remove. Taking into account their respective logs 9 and 11 and their operating profile specific to the region of the world in which they are operated, what is the remaining potential of each truck (in terms of cargo transported, miles traveled, average speed, useful life) and what is the total cost of scheduled maintenance 7 for the remainder of their useful life? What price can the operator ask for for each truck? - An operator has several trucks that have been operated in Senegal up until now and are approaching the end of their useful life. Considering the typical usage profile specific to Senegal and the specific profile in France, taking into account their respective usage and maintenance logs, is it better to keep the trucks operating in Senegal or to redeploy them to France? - For each truck in the operating fleet, taking into account its logs 9 and 11, taking into account its optimum end of life date, taking into account the maximum possible usage until the end of its useful life, taking into account the total costs of future maintenance, when will it no longer be profitable to use the truck in question?

[0287] To answer these questions, the present invention allows optimal life cycle information specific to each truck of the fleet to be refreshed at intended periodicity, or even in real time, thus allowing timely and optimized decisions regarding the management of the assets these trucks represent. Thus, the system 1 allows optimal life cycle information of each piece of equipment 2 of the operator's fleet to be refreshed at intended periodicity, at all times, or even in real time, thus allowing optimized and well-predicted decisions for the benefit of the operators and maintenance technicians.

[0288] The present invention allows for significantly improved design, manufacture, maintenance and usage control of a given series of equipment 2, with long-term visibility into the useful life of each equipment 2. Thus, the present invention allows for the following benefits in terms of control, safety and reliability of usage, as well as economic benefits, for the benefit of designers, manufacturers, maintenance engineers and operators:

[0289] The present invention releases access to control of optimal operation and maintenance decisions for a given device 2 when these decisions involve a short period of time (i.e., they relate to the maintenance 6 to be performed and the amount of usage during the planned period 70).

[0290] On the one hand, the invention makes it possible to determine the optimal maintenance decision for the maintenance 6 to be performed, customized to the complete logs of the equipment 2 (production and maintenance log 9 and usage log 11) and adapted to the intended usage (scenario 8) of the equipment 2 after said maintenance in order to allow zero failures until the date of the next scheduled maintenance 7, pushing back as far as intended the usage limits adapted to the intended zero failures over the planned period 70.

[0291] On the other hand, the present invention eliminates the uncertainty of failures in a controlled way. During operation, the present invention actually provides the operator with real-time refreshed information of the limit usage that meets zero failures until the next maintenance, and the operator then knows the limit to which the usage of the equipment 2 can be pushed without compromising the intended zero failures, or to which the usage must be limited in order to benefit from the intended zero failures. The present invention thus unlocks access to the safety and reliability of use of the equipment 2 and the facility 3.

[0292] Indeed, the present invention is a paradigm shift, by eliminating in a controlled way the uncertainty of failure during operation. The operator no longer needs to know if a failure will occur or if the failure prediction is correct. Now it is the operator who determines not only the date of failure (e.g. the date of the next scheduled maintenance 7) but also the pre-failure usage margin (or the margin between the intended usage of scenario 8 and the limited usage that meets zero failures), and now it is the operator who controls the usage of the equipment 2 to meet the intended zero failures.

[0293] The invention therefore provides a real response that meets the stringent requirements of availability of equipment 2 (or safety of facility 3, if availability of equipment 2 is a prerequisite for the safety of facility 3), and therefore short-term economic advantages are obtained from the invention.

[0294] During the steps of maintenance 6 to be performed, as well as in the course of operation of the equipment 2 during the scheduled period 70, the present invention eliminates the uncertainty of operational failures for the equipment 2 (by accessing optimal maintenance decisions and then by refreshed knowledge of the limit usage amounts that meet zero failures) and does so in a controlled manner.

[0295] On the one hand, the invention therefore allows the entire production time of the facility 3 as intended by the operator. On the other hand, the invention therefore allows a better reliability guarantee with regard to maintenance and operation, and therefore the invention allows the insurance policies of maintenance engineers and operators to be further optimized. The invention also unlocks access to the control of operation and maintenance decisions, including over long periods.

[0296] Indeed, with regard to long-term usage of the device 2, the present invention makes it possible to predict the usage that may be obtained from the device 2 over the long period remaining of its useful life (with refreshed information on the device 2's limited usage and its optimal useful life) and to determine the potential of the device 2 in a timely and constant manner.

[0297] With regard to long-term maintenance of equipment 2, the present invention also makes it possible to timely and at all times predict, over the long term of the useful life of equipment 2 (and for each possible usage profile of the equipment), the industrial tasks for various future maintenance operations, the total cost of future maintenance, the necessary and sufficient stock of spare parts (thus reducing the responsiveness constraints of the supply chain in facing unforeseen needs for spare parts).

[0298] In addition, because the present invention makes it possible to assess and refresh the maximum remaining potential as the total cost of scheduled maintenance 7 remaining during the life of the equipment 2 at any time during the life of the equipment 2, the present invention also makes it possible to determine the remaining value of the equipment at any time during the life of the equipment 2 and to inform the decision to retain or resell the equipment in an associated manner.

[0299] Thus, refreshing optimal life cycle information for equipment 2 or a fleet of equipment with intended periodicity, or even in real time, allows optimized decisions at the right time regarding the management of the assets these equipments represent. The invention also allows for better control of the design and manufacturing choices of equipment 2 of a series, by optimizing the market strategy of the designer / manufacturer and by enabling better reliability assurance regarding the design and manufacturing. This increased control is accompanied by economic benefits due to the optimization that allows the insurance policy of the designer / manufacturer. On the one hand, the invention allows for optimizing the market strategy (market positioning) of the designer / manufacturer.

[0300] For a usage profile associated with a market segment for which equipment 2 is intended, the present invention indeed allows to characterize the optimal life cycle of equipment 2 by generating relevant metrics (margin between target and limit usage, optimal useful life, optimal manufacturing costs, and maintenance schedule over the useful life). These metrics allow to evaluate the suitability of equipment 2 for the segment of interest. Thus, the present invention allows the designer / manufacturer to: Benefit from a market strategy point of view: the designer / manufacturer can thus prioritise targeting of segments for which the device 2 optimises its suitability. - Benefits in terms of market positioning: the designer / manufacturer can therefore better assert the suitability of the equipment 2 to the needs of the segment, in terms of metrics of customers and competitors, support models. On the other hand, the invention also allows for a guarantee of superior reliability in terms of design and manufacturing. The invention indeed allows for determining optimal manufacturing parameters adapted to a given usage profile (for example adapted to a given market segment intersecting with a given geographical segment). The invention also allows for determining possible approaches to optimize the design for a given usage profile. The invention therefore allows for a further optimization of the designer / manufacturer's insurance policy. For this reason, the invention is disruptive in terms of traditional machine learning projects, especially in its models 16, by systematizing the intersection of functional evaluation with data and improving the consideration of functional evaluation in the processing of data. In doing so, the invention is disruptive, especially in terms of the state of the art, and in particular in terms of predictive maintenance techniques, since it is more relevant, more powerful and offers a more general application than simple failure prediction.

Claims

1. A digital system (1) for monitoring the operation and maintenance of at least one industrial device (2) in a facility (3), executed by at least one computing terminal, comprising at least - installing, in a facility (3), at least one piece of industrial equipment (2) resulting from a manufacturing process (4) and representing a series, and then putting said equipment (2) into operation at least for a period (5) until a maintenance step (6) to be carried out; - defining at least one planned maintenance (7) following said maintenance (6) to be carried out after at least one scenario (8) having planned conditions of use (110) of said equipment (2) over a planned operating period (70), The manufacturing (4), installation, operation and maintenance of the equipment (2) comprises at least - a production and maintenance log (9), a task (90) for manufacturing said at least one device (2) until said installation; a production and maintenance log (9) optionally including at least one proactive maintenance (10) task (90) of said equipment (2); a usage log (11) of said equipment (2) over said period (5) between said installation and said maintenance (6) to be carried out, said usage log (11) comprising the conditions of use (110) of said equipment (2) during said period (5); - creating a log (120) of a status (13) of said equipment (2), said log (120) of status including a material indicator (120) of said equipment (2); A technical analysis of the equipment (2) determines at least one correlation (14) between at least one of the tasks (90) and / or at least one of the conditions of use (110) and at least one of the material indicators (120) of the state (13), the correlation (14) establishing at least one link between causes of aging and consequences of aging of the equipment (2), In the correlation (14), said tasks (90) are characterized by tasks identified as critical and / or said conditions of use (110) are characterized by conditions to which said equipment (2) is sensitive and to which it is exposed during operation or at shutdown, said tasks (90) and said conditions (110) of interest affecting said state (13) of said equipment (2); a material condition (13) of said equipment (2) is characterized by said indicator (120) identified as representative of this condition (13) of said equipment (2); Next, in the correlation (14), - measured physical quantities or functions of said measured physical quantities characterizing the conditions of use (110) to which said equipment (2) is sensitive and to which it is exposed during operation or at rest; a measured physical quantity or a function of said measured physical quantity characterizing the material state (13) of the equipment (2) at a given time is determined, Next, recovering and extracting, for the other instruments of said series, data associated with their tasks (90) identified in said correlation (14) and with said physical quantities or functions of physical quantities related to their conditions of use (110) and their material indicators (120) in order to obtain a data set (15); training at least one virtual model (16) based on said dataset (15); During the maintenance (6) of the device (2) to be carried out, at least one value of said tasks (90) of said production and maintenance log (9) and at least one value of said conditions of use (110) of said usage log (11), - at least one value of the tasks (90) of the maintenance to be carried out (6) and the planned conditions of use (110) of the scenario (8) are submitted to the model (16); The monitoring system (1), characterized in that the model (16) generates a planned state (130) of the equipment (2) after the maintenance (6) is performed, and the planned state (130) is compared with a minimum state (17) identified as necessary for the operation of the equipment (2).

2. at least one variation is made to at least one of the values ​​of the tasks (90) of the maintenance (6) to be performed, - when said values ​​are submitted to said model (16), said values ​​of said variations are introduced; A monitoring system (1) as described in claim 1, characterized in that among all said variations, at least one sufficient determination of said maintenance (6) to be carried out is selected for said planned state (130) of said equipment (2) to be equal to or greater than said minimum state (17) at the time of said planned maintenance (7).

3. for a given maintenance decision, the value of at least one of the planned conditions of use (110) of the scenario (8) is modified; - when said values ​​are submitted to said model (16), said values ​​of said corrections and said values ​​of said integrity decisions are introduced; A monitoring system (1) according to claim 1, characterized in that at said time of said scheduled maintenance (7), a limit is calculated for at least one of said scheduled conditions (110) for said scheduled state (130) of said equipment (2) to be equal to said minimum state (17).

4. - when said value is submitted to said model (16), said selected value of said determination of sufficient conservation is introduced; A monitoring system (1) according to claim 2, characterized in that a maximum limit (18) of said scheduled condition (110) is calculated for this sufficient maintenance decision for said scheduled state (130) of said equipment (2) at said time of said scheduled maintenance (7) to be equal to said minimum state (17).

5. A monitoring system (1) according to claim 4, characterized in that for at least one of the planned usage conditions (110) of the scenario (8), a usage margin is determined as the deviation (182) between the corresponding value and the corresponding maximum limit (18).

6. A monitoring system (1) according to claim 5, characterized in that from among said sufficient decisions, an optimal decision is selected as having an acceptable usage margin or as having at least one of said acceptable deviations (182).

7. - a graphical representation in the form of a chart having at least one curve associated with at least a first one of said predetermined conditions (110) as a function of a second one of said predetermined conditions (110); 7. The monitoring system (1) according to claim 6, characterized in that said chart determines said maximum limit (18) of a first predetermined condition (110).

8. in said correlation (14), said production and maintenance log (9) is reduced to a log of critical tasks (90) in the form of at least one list of consecutive values, each of said values ​​of said list characterizing said task (90) of interest in a given maintenance operation; In each list, only permanent values ​​are selected as the values ​​adopted in the last maintenance operation in which the task (90) in question was performed; Monitoring system (1) according to any one of the preceding claims, characterized in that only said persistent values ​​are kept in said log (9) of said critical tasks (90).

9. In the correlation (14), the function of the measured physical quantity of the use conditions (110) is - calculating the time during which said measured physical quantity lies within at least one range of values; and / or a calculation representative of the variation of at least one of said measured physical quantities, and / or A monitoring system (1) according to any one of claims 1 to 7, characterized in that it comprises counting said at least one variation.

10. In the correlation (14), periodically, - said recovery of new data from at least one manufacturer, maintenance technician and / or operator is repeated, - the new data is then extracted to obtain a complete data set (15), A surveillance system (1) according to any one of claims 1 to 7, characterized in that it subsequently updates the training of the model (16) on the basis of the completed data set (15).

11. When the maintenance (6) is performed, following the maintenance (6), the value of at least one of the tasks (90) of the production and maintenance log (9) and the value of at least one task (90) of the maintenance (6) to be performed; At least one value of the usage conditions (110) in the usage log (11) since the maintenance (6) was performed; and the values ​​of the intended conditions of use (11) of the scenario (8) are submitted to the model (16), A monitoring system (1) according to any one of claims 1 to 7, characterized in that the model (16) refreshes the intended state (130) of the equipment (2) and the intended state (130) is compared with a minimum state (17) identified as necessary for the operation of the equipment (2).

12. - it is assumed that at least said sufficient determination of said maintenance (6) to be performed has been performed and that said scenario (8) has been performed up to said planned maintenance (7) which follows said maintenance (6) to be performed; - then performing at least one variation on at least one of said values ​​of said tasks (90) of said scheduled maintenance (7); - once said values ​​have been submitted to said model (16), the values ​​of said variations and of the next scenario (81) to be foreseen for a planned operating period (710) following said planned maintenance (7) are introduced; - among all said variations, at least one sufficient decision for said scheduled maintenance (7) is selected for said scheduled state (130) of said equipment (2) equal to or greater than said minimum state (17) at said time of said maintenance (71) following said scheduled maintenance (7); at least one of said maximum limits (18) and said margin of use associated with said decision of sufficient conservation and said following scenario (81) thus selected is determined, A monitoring system (1) according to claim 5, characterised in that said steps are then repeated recursively after every other scheduled maintenance in the life of said equipment (2).

13. A monitoring system (1) according to claim 6, characterized in that said optimal decisions are selected from at least said sufficient decisions for a corresponding maintenance.

14. A monitoring system (1) according to any one of claims 2 to 7, characterized in that when a maintenance decision is insufficient at a planned state (130) which is less than said minimum state (17), a failure date (19) is determined for the corresponding maintenance decision.

15. A monitoring system (1) as described in claim 14, characterized in that for the maintenance (6) to be performed or for planned maintenance (7, 71) for which no maintenance decision is identified as sufficient and for at least one given insufficient decision of the maintenance (6) to be performed, the end date of the useful life of the equipment (2) is determined to be the failure date (19) associated with the maintenance decision.

16. for said maintenance (5) to be carried out or for a scheduled maintenance identified as the last maintenance in the life of said equipment (2), 16. The monitoring system (1) according to claim 15, characterized in that the maintenance decision that optimizes any combination between the failure date (19), end-of-life margin and constraints of the task (90) of the last maintenance is selected from among possible maintenance decisions.

17. For at least two dummy devices of the same series of the device (2) associated with separate production determinations, a simulation is carried out by submitting to said model (16) said at least two manufacturing decisions and at least one planned usage scenario (8) over said expected useful life of said two dummy devices; said model (16) generates at least one predefined state (130) for each of said two dummy machines; one of said at least two manufacturing decisions is selected depending on said predetermined states (130) of said two dummy devices; A monitoring system (1) according to claim 1, characterised in that the selected manufacturing decisions are accessible to the designer / manufacturer.

18. for at least two dummy devices of the same series of said device (2) associated with two separate production determinations, a recursive simulation is likewise performed for each of said at least two production decisions to determine the optimal life cycle associated with each of said production decisions, i.e. the optimal maintenance decisions of a series, the maximum limits (18) and the usage margins of the series associated with these optimal maintenance decisions, and the associated optimal useful life of the dummy equipment; - optimal manufacturing decisions are selected depending on the results of said simulation, A monitoring system (1) according to claim 12, characterized in that said selected optimal manufacturing decisions are accessible to said designer / manufacturer.

19. - applied to a fleet of several machines (2) of said series belonging to a single operator, the results obtained are combined for each of said devices (2), A monitoring system (1) according to any one of claims 1 to 7, characterized in that said results are accessible to at least said operator.

20. A surveillance system (1) according to any one of claims 1 to 7, characterized in that the training of the model (16) belongs to the field of artificial intelligence and can be machine learning.

Images