Dynamic digital replicas of operational facilities

Abstract

A method for generating a dynamic digital replica includes aggregating separate classes of facility data pertaining to one or more of a plurality of equipment of a operational facility to thereby produce aggregated data, performing qualitative analysis on the aggregated data to extract one or more features associated with one or more of the plurality of equipment, wherein each of the extracted features is based on a plurality of the separate classes of facility data, contextualizing the aggregated data and the extracted features using unstructured data that is separate from the aggregated data to generate contextualized data, generating a dynamic digital replica of the operational facility based on the aggregated data, the extracted features, and the contextualized data, and providing, using the dynamic digital replica, a recommendation to a user pertaining to an action to be performed on at least one of the plurality of equipment of the operational facility.

Description

DYNAMIC DIGITAL REPLICAS OF OPERATIONAL FACILITIES CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] Not applicable. STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0002] Not applicable. BACKGROUND

[0003] Operational facilities, such as offshore platforms, well systems, chemical and refining plants, manufacturing plants, power stations, hydrogen production facilities, renewable energy facilities such as solar arrays and wind farms, and transportation hubs, rely on complex systems of machinery and infrastructure. Traditional methods for monitoring and managing these facilities often involve time-consuming and costly physical inspections. Digital replica technology aims to replicate these facilities in a virtual environment, allowing for real-time monitoring, predictive maintenance, and simulation. BRIEF SUMMARY OF THE DISCLOSURE

[0004] An embodiment of a computer-implemented method for generating a dynamic digital replica of an operational facility comprises (a) aggregating separate classes of facility data each pertaining to one or more of a plurality of equipment of the operational facility to thereby produce aggregated data, (b) performing qualitative analysis on the aggregated data to extract one or more features associated with one or more of the plurality of equipment, wherein each of the extracted features is based on a plurality of the separate classes of facility data, (c) contextualizing the aggregated data and the extracted features using unstructured data that is separate from the aggregated data to generate contextualized data, (d) generating a dynamic digital replica of the operational facility based on the aggregated data, the extracted features, and the contextualized data, and (e) providing, using the dynamic digital replica, a recommendation to a user pertaining to an action to be performed on at least one of the plurality of equipment of the operational facility. In some embodiments, the method comprises (f) determining, using the dynamicdigital replica, whether the operational facility is in a known state or an unknown state prior to (e), and (g) providing, using the dynamic digital replica, the recommendation to the user at (e) in response to determining at (f) that the operational facility is in the known state. In some embodiments, the method comprises (f) determining, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e), and (g) requesting, using the dynamic digital replica, a user for input pertaining to the operational facility following a determination at (f) that the operational facility is in the unknown state. In certain embodiments, the method comprises (h) receiving as feedback data, using the dynamic digital replica, the requested input pertaining to the operational facility from the user to which input is requested at (g), (i) determining, using the dynamic digital replica, that the operational facility is in the known state in response to receiving the feedback data, and (j) providing the recommendation to the user at (e) in response to (i). In certain embodiments, the method comprises (f) applying, using the dynamic digital replica, feedback data from a user to the aggregated data, wherein the feedback data is based on the recommendation made to the user at (e). In some embodiments, (f) comprises using principal component analysis (PCA) to update a dimensionality of the aggregated data based on the feedback data. In some embodiments, the method comprises (f) determining, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e), and (g) searching, using the dynamic digital replica, one or more of the aggregated data, the extracted features, and the contextual data to generate additional contextualized data following a determination at (f) that the operational facility is in the unknown state. In certain embodiments, the method comprises (h) determining, using the dynamic digital replica, that the operational facility is in the known state in response to producing the additional contextualized data at (g), and (i) providing the recommendation to the user at (e) in response to (h). In certain embodiments, the method comprises (f) determining, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e), (g) searching, using the dynamic digital replica, one or more of the aggregated data, the extracted features, and the contextual data to generate additional contextualized data following a determination at (f) that the operational facility is in the unknown state, (h) determining, using the dynamic digital replica, that the operational facility remains in the unknown state following (g), and (i) requesting a from for a user input pertaining to theoperational facility following the determination at (h) that the operational facility remains in the unknown state. In some embodiments, the method comprises (j) receiving, by the dynamic digital replica, as feedback data the requested input pertaining to the operational facility from the user from which input is requested at (i), (k) determining, using the dynamic digital replica, that the operational facility is in the known state in response to receiving the feedback data, and (l) providing the recommendation to the user at (e) in response to (k). In some embodiments, (c) comprises perform a principal component analysis (PCA) on the aggregated data to reduce a dimensionality of the aggregated data to generate the contextualized data.

[0005] An embodiment of a computer-readable medium storing executable code which, when executed by a processor, causes the processor to (a) aggregate separate classes of facility data each pertaining to one or more of a plurality of equipment of an operational facility to thereby produce aggregated data, (b) perform qualitative analysis on the aggregated data to extract one or more features associated with one or more of the plurality of equipment, wherein each of the extracted features is based on a plurality of the separate classes of facility data, (c) contextualize the aggregated data and the extracted features using unstructured data that is separate from the aggregated data to generate contextualized data, (d) generate a dynamic digital replica of the operational facility based on the aggregated data, the extracted features, and the contextualized data, and (e) provide, using the dynamic digital replica, a recommendation to a user pertaining to an action to be performed on at least one of the plurality of equipment of the operational facility. In certain embodiments, when executed by the processor, causes the processor to (f) determine, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e), and (g) provide, using the dynamic digital replica, the recommendation to the user at (e) in response to determining at (f) that the operational facility is in the known state. In certain embodiments, when executed by the processor, causes the processor to (f) determine, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e), and (g) request from a user input pertaining to the operational facility following a determination at (f) that the operational facility is in the unknown state. In some embodiments, when executed by the processor, causes the processor to (h) receive, by the dynamic digital replace, as feedback data the requested input pertaining to the operational facility from the user fromwhich input is requested at (g), (i) determine, using the dynamic digital replica, that the operational facility is in the known state in response to receiving the feedback data, and (j) provide the recommendation to the user at (e) in response to (i). In some embodiments, when executed by the processor, causes the processor to (f) apply, using the dynamic digital replica, feedback data from a user to the aggregated data, wherein the feedback data is based on the recommendation made to the user at (e). In certain embodiments, (f) comprises using principal component analysis (PCA) to update a dimensionality of the aggregated data based on the feedback data. In certain embodiments, when executed by the processor, causes the processor to (f) determine, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e), and (g) search, using the dynamic digital replica, one or more of the aggregated data, the extracted features, and the contextual data to generate additional contextualized data following a determination at (f) that the operational facility is in the unknown state. In some embodiments, when executed by the processor, causes the processor to (h) determine, using the dynamic digital replica, that the operational facility is in the known state in response to producing the additional contextualized data at (g), and (i) provide the recommendation to the user at (e) in response to (h). In some embodiments, when executed by the processor, causes the processor to (f) determine, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e), (g) search, using the dynamic digital replica, one or more of the aggregated data, the extracted features, and the contextual data to generate additional contextualized data following a determination at (f) that the operational facility is in the unknown state, (h) determine, using the dynamic digital replica, that the operational facility remains in the unknown state following (g), and (i) request a from a user input pertaining to the operational facility following the determination at (h) that the operational facility remains in the unknown state.

[0006] Embodiments described herein comprise a combination of features and characteristics intended to address various shortcomings associated with certain prior devices, systems, and methods. The foregoing has outlined rather broadly the features and technical characteristics of the disclosed embodiments in order that the detailed description that follows may be better understood. The various characteristics and features described above, as well as others, will be readily apparent to those skilled in the art uponreading the following detailed description, and by referring to the accompanying drawings. It should be appreciated that the conception and the specific embodiments disclosed may be readily utilized as a basis for modifying or designing other structures for carrying out the same purposes as the disclosed embodiments. It should also be realized that such equivalent constructions do not depart from the spirit and scope of the principles disclosed herein. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] For a detailed description of exemplary embodiments of the disclosure, reference will now be made to the accompanying drawings in which:

[0008] FIG.1 depicts a computer system for displaying a screen image showing a portion of a basic digital replica of an operational facility according to some embodiments;

[0009] FIG. 2 depicts a graphical user interface (GUI) showing a view of the operational facility according to some embodiments;

[0010] FIG. 3 depicts a GUI showing another view of the operational facility according to some embodiments;

[0011] FIG.4 depicts a block diagram of a dynamic digital replica of an operational facility according to some embodiments;

[0012] FIG.5 depicts a graph presenting time series data according to some embodiments;

[0013] FIG.6 depicts a diagram of unstructured data according to some embodiments;

[0014] FIG.7 depicts a knowledge graph according to some embodiments;

[0015] FIG. 8 depicts a block diagram of an aggregation process according to some embodiments;

[0016] FIG.9 depicts a block diagram of a qualitative analysis process according to some embodiments;

[0017] FIG.10 is a schematic depiction of a discrete Fourier transform according to some embodiments;

[0018] FIG.11 depicts a block diagram of a bowtie model according to some embodiments;

[0019] FIG.12 depicts a block diagram of another dynamic digital replica of an operational facility according to some embodiments;

[0020] FIG. 13 depicts a block diagram of a computer system according to some embodiments; and

[0021] FIG. 14 depicts a flowchart of a computer-implemented method for generating a dynamic digital replica of an operational facility is shown according to some embodiments. DETAILED DESCRIPTION OF THE DISCLOSED EMBODIMENTS

[0022] The following discussion is directed to various exemplary embodiments. However, one skilled in the art will understand that the examples disclosed herein have broad application, and that the discussion of any embodiment is meant only to be exemplary of that embodiment, and not intended to suggest that the scope of the disclosure, including the claims, is limited to that embodiment.

[0023] Certain terms are used throughout the following description and claims to refer to particular features or components. As one skilled in the art will appreciate, different persons may refer to the same feature or component by different names. This document does not intend to distinguish between components or features that differ in name but not function. The drawing figures are not necessarily to scale. Certain features and components herein may be shown exaggerated in scale or in somewhat schematic form and some details of conventional elements may not be shown in interest of clarity and conciseness.

[0024] In the following discussion and in the claims, the terms "including" and "comprising" are used in an open-ended fashion, and thus should be interpreted to mean "including, but not limited to…” Also, the term "couple" or "couples" is intended to mean either an indirect or direct connection. Thus, if a first device couples to a second device, that connection may be through a direct connection, or through an indirect connection via other devices, components, and connections. In addition, as used herein, the terms "axial" and "axially" generally mean along or parallel to a central axis (e.g., central axis of a body or a port), while the terms "radial" and "radially" generally mean perpendicular to the central axis. For instance, an axial distance refers to a distance measured along or parallel to the central axis, and a radial distance means a distance measured perpendicular to the central axis.

[0025] As described above, digital replica technology aims to replicate operational facilities in a virtual environment, allowing for real-time monitoring, predictive maintenance, and simulation. Using offshore production facilities as an example, some offshore production facilities are referred to as floating cities because even the smallest offshore platforms have more than a dozen people living and working aboard. Larger facilities, which may beanchored to the sea bed below, might have as many as 200 people aboard to operate the facility effectively. Maintaining production facilities that pump thousands of barrels of oil a day from a subterranean reservoir disposed below the sea floor utilizes—at all times—a variety of essential personnel performing different jobs including, without limitation, operators, maintenance technicians, welders, divers, engineers, cooks, safety personnel, and medical personnel. In addition to such personnel, additional maintenance personnel regularly travel to the facility for equipment inspections and maintenance. At least some of these regular maintenance activities are preventive in nature, and thus, are performed visually.

[0026] In addition to the regular maintenance activities, an equipment survey crew is typically sent aboard when the offshore production facility to inspect pipework for corrosion or when equipment is modified, for example, by replacing an important piece of equipment, such as a pump. These multiple layers of inspection and maintenance add to the overall cost, are time consuming, and create redundancies. In addition, offshore production facilities are inherently complex and can be dangerous; the constant influx and outflow of people may add to risks. Thus, it is generally desirable to reduce the number of people aboard an offshore production facility at any given time.

[0027] To improve the quality, performance, and efficiency of inspection and maintenance, the power of automation and data is used by the facility operators, engineers, maintenance and inspection crews to aid their decision making. In some cases, the offshore production facility may use the following types of maintenance-related software tools: integration and data management software, operations planning and reporting software, analytics and assurance software, real-time solutions software, equipment condition monitoring software, etc. The list of software tools mentioned herein is not exhaustive. Each of the above- mentioned software tools performs a multiplicity of different functions and generates a surfeit of distributed data relevant to its operators. The use of a large number of software tools is sometimes cost ineffective and generates distributed data, which may create redundancies and reduce the overall efficiency of the inspection or maintenance of the facilities. Thus, there is a need for systems and methods to mitigate the issues mentioned above. In particular, there is a need to aggregate all the distributed data to reduce redundancies and fully utilize the generated data. Such systems and methods mayimprove the quality of equipment maintenance, reduce the number of people aboard at any given time, and enhance the use of automation tools overall.

[0028] Digital replica technology aims to address at least some of these issues by replicating the features of a given operational facility in a virtual environment. However, conventional digital replicas of operational facilities suffer from several limitations which limit their efficacy and applicability. For example, at least some conventional digital replicas may struggle effectively integrating diverse datasets from various sources data sources such as, for example sensors, internet-of-things (IoT) devices, and legacy systems. In addition, at least some conventional digital replicas are limited in their abilities to accurately reflect the current or real-time state of the operational facility and / or are limited in their ability to forecast future states of the operational facility whereby proactive maintenance and optimization of the operational facility may be implemented. Further, at least some conventional digital replicas are limited in their scalability such that digital replicas may be created and deployed for facilities of various scales or configurations without requiring significant customization.

[0029] Accordingly, embodiments of dynamic digital replicas are presented herein that address the limitations outlined above. In some embodiments, a dynamic digital replica may be generated by aggregating separate classes of facility data each pertaining to one or more of a plurality of equipment of the operational facility to thereby produce aggregated data. This includes automatically aggregating both structured and unstructured data as will be described further herein. The dynamic digital replica may be generated by additionally performing qualitative analysis on the aggregated data to extract one or more features associated with one or more of the plurality of equipment, wherein each of the extracted features is based on a plurality of the separate classes of facility data. This may include extracting images or text from unstructured data such as composite data including different types of features (e.g., images with embedded text) that must be separately extracted.

[0030] The dynamic digital replica may be generated by additionally contextualizing the aggregated data and the extracted features using unstructured data that is separate from the aggregated data to generate contextualized data. In this manner, the dynamic digital replica may automatically uncover relationships between different features of the aggregated data that may be useful in explaining the occurrence of issues or scenarios which may arise with the operational facility.

[0031] The aggregated data may form or define a basic digital replica of the operational facility. As will be discussed further herein, a dynamic digital twin of the operational facility may be generated from an underlying basic digital twin (comprising aggregated classes of data) through the application of additional techniques including, among other things, contextualizing the aggregated data and extracted features using unstructured data that is separate from the aggregated data to generate contextualized data. In some embodiments, a dynamic digital replica of the operational facility may be generated based on the aggregated data, the extracted features, and the contextualized data. In some embodiments, the generated dynamic digital replica may provide one or more recommendations to a user pertaining to an action to be performed on at least one of the plurality of equipment of the operational facility.

[0032] Initially, techniques for generating a basic digital replica of an operational facility will be discussed, where embodiments of dynamic digital replicas may be generated from the underlying basic digital twin, as will be discussed further herein. Refer now to FIG. 1, a computer system 100 for displaying a screen image 101 illustrating a portion of a basic digital replica 102 of an operational facility is shown. In this exemplary embodiment, the operational facility comprises an offshore production facility; however, it may be understood that the type of operational facility may vary depending on the given application. For example, in other embodiments, the operational facility may comprise a subterranean wellbore drilling or completion facility, a refining, chemical, or other fluid processing facility, a construction facility, a hydrogen production facility, renewable energy facilities such as solar arrays and wind farms, and so on and so forth.

[0033] The computer system 100 may be the computer system discussed below with respect to FIG.12, for example. The screen image 101 of FIG.1 illustrates the portion of the offshore production facility positioned above the sea. The screen image 101 includes a platform of the offshore production facility built on concrete or steel legs anchored directly onto the seabed. The additional equipment that is present on the seabed, for example, trees, manifold, pipeline end termination system (PLET), riser, jumper, flowlines, etc., are not shown in FIG.1. However, it should be appreciated that the description herein relates to digital replica of the complete offshore production facility (in this example, the offshore production facility being other types of operational facilities in other examples), not just the portion of the basic digital replica 102 shown in the screen image 101.

[0034] The basic digital replica 102 is a computer-generated visualization of the complete offshore production facility such that the basic digital replica 102 acts as an information databank in all four dimensions of time and space. In particular, the basic digital replica 102 transforms currently implemented software systems that generate distributed dataset into a four-dimensional (three-dimensional (3D) space plus time) repository that can be visually and temporally browsed and analyzed. The time domain may be provided by the sensors placed on the production facility. For example, the sensors may be positioned on all the equipment used in the offshore production facility, where the sensors track the health of the equipment and provide dynamic tracking reports. These tracking reports may be accessed by the operator and may also be used to visually indicate the health of the equipment.

[0035] FIG. 1 also depicts various illustrative attributes of the basic digital replica 102, some of which enable the basic digital replica 102 to function as an interactive rendering of the offshore production facility. It should be appreciated that the attributes shown in FIG.1 are not an exhaustive list. The attributes are shown to illustrate some of the capabilities of the basic digital replica 102. Given the wide variation of equipment, conditions, and situations that may be encountered in the offshore production facility, all possible attributes that may be needed for all the possible scenarios cannot possibly be presented in this specification. As such, only some examples of such scenarios will be provided in this specification.

[0036] Illustrative attributes of the basic digital replica 102 will now be described. One of the key attributes of the basic digital replica 102 is a graphical user interface (GUI) 130 that is accessible through a two-dimensional (2D) pixel matrix of display unit, for example, monitors of computers or display screens of other electronic devices, such as mobile phones, handheld computers, or computer-interfaced image projection devices, or a combination thereof. The GUI 130 interactively displays a four-dimensional (4D) view of the offshore production facility onto the 2D pixel matrix of display screens. The GUI 130 may be a component located on the production facility, for example. The GUI 130 provides the operator with a set of widgets, such as buttons, sliders, choice and list boxes, which enable the operator to generate requests, which may include transferring from one type of viewpoint to another. The GUI 130 may be accessible utilizing a web browser that provides the operator with access to the most recent version of the GUI 130, independent of thecombination of hardware and operating system utilized and independent of the location of the hardware and operating system. For example, the basic digital replica 102 may be viewed at a location remote from the actual location of the offshore production facility, such as a regional service office or an operations headquarters.

[0037] Referring briefly to FIG.2, an illustrative GUI 200 is depicted. The GUI 200 may be the GUI 130 (FIG. 1), for example. The GUI 200 includes a basic digital replica of the complete offshore production facility, which includes the view of the basic digital replica 102 as shown in FIG.1 and other components 208 that may form the offshore production facility. The components 208 are shown to be present on a sea bed. The GUI 200 also depicts a map 205 that shows the complete offshore production facility. The GUI 200 further depicts a set of widgets 212 that are present on the top and side of the GUI 200. In one example, the widgets 212 on the top of the GUI 200 may include a tree view option and a change view option, and the widgets 212 on the side of the GUI 200 may include a filter option, a dimension measurement option, an adding text option, a changing color option, etc.

[0038] Assume that the GUI 200, when turned on for the first time, displays a full-field digital replica view (as shown in FIG.2) illustrating a common operating view 138 (FIG.1), which includes a real-time weather view 210 and a view of locations of vessels 214 overlaid upon a view of all other components coupled to the offshore production facility and sharing similar map coordinates to the real-time weather view 210 and the view of the locations of vessels 214. The view of locations of vessels 214 may be generated utilizing an automatic identification system (AIS). The AIS shows real-time locations of the vessels using transponders on the ships via a vessel tracking 144 (FIG.1). Now, the operator can select the appropriate widgets to access a desired basic digital replica, for example, the offshore production facility as shown in the view of the basic digital replica 102. After selecting the offshore production facility as represented by the view of the basic digital replica 102 in the GUI 200, the operator may see a view as shown in screen image 101 as shown in FIG.1, for example.

[0039] Referring again to FIG. 1, as noted above, the basic digital replica 102 of the offshore production facility is interactive in that operators can manipulate the spatial location and orientation of the perspective viewpoint of the facility via a user interface device 131, such as a mouse, joystick, trackball, or touchscreen, to create an effect ofactually walking through the computer-generated 4D visual scene. This walking through attribute is herein referred to as a digital walkthrough 136. During the digital walkthrough 136, the operator may perform a component ID 137 by selecting, highlighting, identifying, and accessing various digital equivalents of the equipment. This creates a first-person user experience of the facility even when being remote. To perform a digital walkthrough 136, for example, the operator may zoom in and select to view one of the digital walkways 103 on the facility. While on one of the digital walkways 103, the operator may digitally walk around the platform and, using the user interface device 131, select a piece of equipment in the instant visual scene. For example, the operator may select a digital equivalent of a piece of equipment, an iron roughneck present in the instant visual scene, of one of the digital walkways 103. In other examples, the operator may utilize the component ID 137 to view the component without first selecting one of the digital walkways 103.

[0040] As can be seen in FIG.1, some of the attributes are depicted as being linked with other attributes. However, such links between attributes does not imply a relationship, such as dependence or causality. The links are merely shown to provide examples of the way in which one or more of the attributes can function together. There may be one or more of the attributes that may not be linked but that can function together. For example, an operator using the user interface 130 can perform the digital walkthrough 136 and can perform the component ID 137 during the digital walkthrough 136. In some examples, equipment can be identified by digital tags 115 that may mark each piece of equipment present in the digital walkthrough. In other examples, the operator may need to utilize the user interface device 131 to select the digital equivalent of the equipment to see the digital tags 115.

[0041] In addition to identifying components, the basic digital replica 102 is configured to provide more details on the identified component. For the sake of illustration, assume that the operator identifies a digital component 109 as a pump while digitally walking through one of the digital walkways 103 of the basic digital replica 102. The operator can access real-time operation metrics 105 of the identified digital component 109. For example, the operator can gather live equipment and process data such as vibration data, shaft speed, flow rate, pressure, and temperature, etc. of the pump. In addition to real-time operation metrics 105, the operator can also access reports 119 of the identified digital component109. For example, the operator may access specification reports, schematics, reliability reports, maintenance history, integrity database, data sheet, pump curve, and equipment health assessment, etc. of the pump. After accessing these reports, the operator may review and inspect the design of the identified digital component 109 utilizing a design review and inspection 113.

[0042] In addition to being able to check real-time operation metrics 105, the basic digital replica 102 is configured to perform additional analysis 111 on the instant visual scene or any selected component or piece of equipment. For example, assume that the GUI 130 displays the offshore production platform shown in the screen image 101. The operator, using the appropriate widgets on the GUI, can view a corrosion circuit model of the complete offshore production platform. Corrosion circuit modeling is carried out as part of a risk-based analysis (RBA). Corrosion circuit modeling combines fluid type and piping materials or chemical make-up into systems or sub-systems, which can be grouped into corrosion or erosion mechanisms. These mechanisms are monitored over the operating lifetime of the facility in accordance with an operating management system (OMS) that defines a systematic and consistent approach for managing operating activities. The OMS may be implemented as an OMS application of other applications 140 available through the GUI 130. By utilizing the OMS, the basic digital replica 102 allows visualization and connection of disparate data to improve performance, reinforcing a commitment to operate safe and reliable operations compliant with the OMS. This monitoring is further utilized in the integrity management plan (IMP) which, in some examples, may form a part of the overall asset integrity management system. In addition to viewing the corrosion circuit of the offshore production platform, the operator may isolate a particular corrosion circuit and access risk-based analysis reports of the isolated corrosion circuit. Furthermore, the operator may also access piping and instrumentation diagram of the isolated corrosion circuit.

[0043] The additional analysis is not limited to corrosion circuits. In other examples, additional analysis 111 may also include comparing theoretical design calculations and actual operating conditions – for example comparing actual erosion data with erosion modeling data, actual turbine thermodynamic data with turbine thermodynamic modeling, etc. These comparisons are monitored over the operating lifetime of the facility in accordance with the OMS. In some examples, the basic digital replica may allow viewingof a predicted aging of a piece of equipment, of a system, or of the production facility as a whole. The predicted aging may be based on one of the plurality of models available. For example, the basic digital replica 102 may provide a view of corrosion within the production facility six months from a current date utilizing theoretical calculations, actual operating conditions, past data, or a model based on a combination thereof to predict and display the aging of the piece of equipment, the system, or of the production facility as a whole. This capability allows an operator to schedule maintenance as well as predict possible operations interruptions for further analysis.

[0044] The GUI 130 may allow access to a map 134 that provides the operator another spatial viewpoint on the operator’s instant spatial location on or in the basic digital replica 102. For example, using the user interface device 131, the operator may digitally walk to a location in the basic digital replica 102 where a pump is placed. At this point, the operator may see on the display screen her current location within the basic digital replica 102. The current location in the map 134, in some examples, may be labeled or marked as a colored dot.

[0045] The basic digital replica 102 also includes a process surveillance system 107 that provides live process data for selected pieces of equipment or systems of the offshore production facility. Refer briefly to FIG. 3, an illustrative GUI 300 showing a view of equipment 310 placed on the seabed. The GUI 300 may be the GUI 130 (FIG. 1), for example. The equipment 310 may be a sub-system or system shown by selecting one or more than one of the components 208 of the common operating view 138 of the GUI 200, for example. The GUI 300 shows the widgets 212, an overall field view 315, a detail schematic diagram 320, and live process data of the equipment 310. The equipment 310 includes motors and pipes pumping oil in real-time from the reservoir underneath the seabed. The live process data includes flow direction, flow rate, pressure, and temperature. The live process data also allows for the operator to manipulate a sub- system or sub-component of the equipment 310 and analyze the effect it may have on the overall system. For example, the operator, using the user interface device, may turn-off a valve and check its effect on the equipment 310. This capability is especially useful in examples where a portion of equipment needs to be replaced and the operator wants to check the effect the downtime will have on the other systems. In another example, theoperator may utilize the turned-off valve as an input into the OMS application to determine the effect the downtime will have on the other systems.

[0046] While troubleshooting a problem or scenario, the operator may utilize different attributes of the basic digital replica 102. For example, the operator may access a work order history of the valve from a work order system 117 to review all the preventative and corrective maintenance performed on the valve in the past year. In addition, the operator may also access the real-time operation metrics 105 of the equipment in which the valve is present. The operator can also view the pneumatic schematic in the reports 119. If, after troubleshooting, it is concluded that the valve needs to be replaced, the operator can access a spare parts inventory and confirm whether a warehouse has a replacement part.

[0047] In summary, the basic digital replica 102 facilitates maintenance activities by aggregating all relevant data in one visual repository (basic digital replica 102), where an operator can access the relevant data more efficiently and without creating redundancies.

[0048] In one example, basic digital replica 102 utilizes a digital tag algorithm that facilitates accessing relevant details based on the digital tags 115. The digital tags 115 may be used to identify pieces of the equipment and their associated activities or documents. The digital tag algorithm searches and cross-references all systems and data to generate relevant output based on the digital tags 115 of the identified equipment. This allows users to navigate around the plethora of information visually using the basic digital replica 102. Additionally, using the digital tag algorithm reveals discrepancies between data sources and allows rectification of the discrepancy at its source, which in turn improves quality and accuracy. By implementing the digital tag algorithm in compliance with the OMS, a systematic and consistent approach for managing data is provided.

[0049] In some examples, the user interface 130 may also provide access to other applications 140. The other applications 140 are machine-readable instructions that cause a processor to perform the actions specified or to cause the actions to be performed by another component of a computer system. The computer system may be the computer system 100, for example. The other applications 140 include Geospatial Information System (GIS) 142, and vessel tracking 144, for example. GIS 142 is configured to provide region-wide visibility of vessels, facility, subsea structure, and reservoir development data. Integrating GIS 142 with the basic digital replica 102 assists with the vessel tracking 144. GIS 142 provides details related to the offshore production facility, including coordinates ofthe equipment, vessels, production facility, etc., and assists with defining a coordinate- based view of the common operating view 138, for example.

[0050] In some examples, the basic digital replica 102 is configured to provide equipment training competency assessments to essential personnel. The basic digital replica 102, in some examples, may be configured to perform simulations 156. The simulations 156 may be material handling simulations, for instance. In one example, the simulations 156 may be accessed by selecting a digital walkway. In another example, the simulations 156 may be accessed by selecting a digital component or a digital system. In yet another example, the simulations 156 may be accessed utilizing the GUI 130. In some examples, the GUI 130 may be assumed by a mixed-reality device 125. The mixed-reality device 125 may also be a component of the production facility, for example. The GUI 130 may display on a display screen of the mixed-reality device 125. The functions an operator may perform using the GUI 130 applies to the mixed-reality device 125, but instead of using the user interface device 131, the operator may use a user interface device related to the mixed- reality headset. The user interface device related to the mixed-reality device 125 may be sensors for detecting hand movements, for example.

[0051] Referring now to FIG. 4, a block diagram of a dynamic digital replica 400 of an operational facility is shown according to some embodiments. Dynamic digital replica 400 includes a plurality of modules encoding instructions executable by a computer system for creating and / or operating by a user the dynamic digital replica 400.

[0052] The dynamic digital replica 400 receives a variety of different classes of data from a range of data sources associated with a given operational facility. In this exemplary embodiment, dynamic digital replica 400 ingests separate classes of facility data 390-394 each pertaining to one or more of a plurality of equipment of an operational facility. In some embodiments, once aggregated, the classes of facility data 390-394 may form or define a basic digital replica (e.g., basic digital replica 102 shown in FIG. 1) of the operational facility. The separate classes of facility data 390-394 may comprise structured and / or unstructured data. Particularly, dynamic digital replica 400 may digest one or more three-dimensional (3D) models 390 of the operational facility (or selected components thereof) produced from solid or 3D modeling software. In addition, dynamic digital replica 400 may ingest equipment facility data 391 specifying one or more attributes or features(e.g., design parameters, operational parameters) of various equipment of the operational facility.

[0053] Dynamic digital replica 400 may additionally ingest one or more process diagrams 392 of the operational facility illustrating schematically the layout of different systems (e.g., fluid systems, electrical systems) of interconnected equipment of the operational facility. Dynamic digital replica 400 may additionally ingest one or more construction drawings 393 of the operational facility illustrating where different components (e.g., buildings, fixed equipment) of the operational facility are located in physical space. Further, dynamic digital replica 400 may ingest one or more surface scans 394 of the operational facility. The surface scans 394 may be produced via scanning the operational facility with one or more lasers as part of a laser scanning process, and / or through photographing the operational facility as part of a photogrammetry process. The surface scans 394 may indicate where equipment of the operational facility is physically located along with the size and / or shape of the equipment.

[0054] Referring briefly to FIG.5, a graph 450 presenting time series data including a flow rate datastream 451, a pressure datastream 452, and a temperature datastream 453 pertaining to one or more pieces of equipment of an operational facility. Graph 450 represents a form of structured data ingestible by the dynamic digital twin 400. The datastreams 451-453 captured in graph 450 is structured data where each datastream 451-453 share an association with the operational facility. For example, the datastreams 451-453 may be associated with the same piece of equipment (e.g., the same pump, pressure vessel, heat exchanger) or the same physical location of the operational facility. As used herein, the term “structured data” refers to data that is organized according to a predefined format. Structured data may be stored in databases or tables, and is generally characterized by a clear and predetermined data schema with distinct fields, data types, and relationships, making it easy to search, analyze, aggregate, and otherwise process using various data management tools and techniques.

[0055] Table 1 presents an example of unstructured data ingestible by the dynamic digital replica 400. As used herein, the term “unstructured data” refers to data that, unlike structured data, is not organized according to a predefined format. Unstructured data may include, for example, text documents, images, videos, social media posts, and more. Unstructured data generally does not follow a predefined schema, making it more challenging to aggregate, analyze, and extract insights from using traditional databases or data processing methods. Instead, complex natural language processing and machine learning techniques are often used to make sense of unstructured data.

[0056] In the example presented in Table 1, coordinate-based data is presented in tabular format whereby a given piece of equipment (identified by the “equipment_id” column) is related to a given point id, a set of X, Y, Z coordinates, and an equipment name. Some unstructured data, such as the coordinate-based data presented in Table 1, may contain data that is indirectly rather than directly related. As an example, the datastreams 451-453 may be associated with a particular piece of equipment (e.g., they may each be related to the same fluid conduit), but the sensors responsible for generating datastreams 451-453 (e.g., flow, pressure, and temperature sensors) may not have any readily availablerelationship to the given piece of equipment (e.g., the given fluid conduit) other than geographic proximity decipherable from coordinate-based data like the data presented in Table 1.TABLE 2

[0057] Table 2 presents an additional example of unstructured data ingestible by the dynamic digital replica 400. Particularly, Table 2 includes unstructured data in the form of data sheet-based data. In this example, the unstructured data presented in Table 2 presents design condition parameters of equipment of an operational facility. The unstructured data of Table 2 includes several different fields some of which may be relevant for the operation of dynamic digital replica 400 while others may be irrelevant. In addition, some of the fields contained in Table 2 may be empty and the information presented in Table 2 may change over time. As will be discussed further herein, dynamic digital replica 400 may, when ingesting unstructured data such as the data presented in Table 2, search this type of data (e.g., search the various fields of the data sheet-based data) and aggregate the unstructured data with structured data such as the time series data presented in FIG.5.

[0058] Referring briefly to FIG.6, a further example of unstructured data 470 ingestible by the dynamic digital replica 400 is presented. Particularly, unstructured data 470 comprises composite data (e.g., documents or files containing multiple separate classes of data) in the form of narrative-based data (and thus may also be referred to herein as narrative-based data 470). In this example, narrative-based data 470 presents laydown areas and material handling routes overlaid on a two-dimensional (2D) layout of a deck of an operational facility. Particularly, narrative-based data 470 includes both a textual narrative 472 and a layout or map 474 directly related to the textual narrative 472. As will be discussed further herein, dynamic digital replica 400 may, when ingesting composite unstructured data such as the narrative-based data 470 presented in FIG.6, read the text provided therewith (e.g., textual narrative 472) and utilize coordinates and other information contained in the composite data (e.g., information contained in the map 474) to aggregate the composite unstructured data with other classes or forms of data. As an example, dynamic digital twin 400 may aggregate the narrative-based data 470 of FIG. 6 with additional data such as weight data to, for instance, calculate deck loading during movement and storage of equipment of the operational facility.

[0059] In this exemplary embodiment, dynamic digital replica 400 includes an aggregation engine or module 405 that receives or ingests the various classes of facility data 390-394 outlined above. In other embodiments, the classes of data ingested by aggregation module 405 may vary from that shown in FIG.4, where facility data 390-394 is only meant to be exemplary in nature. Aggregation module 405 is generally configured to aggregate the various classes of facility data 390-394 ingested by dynamic digital replica 400 to produce aggregated data 406. In some embodiments, this includes aggregating by the aggregation module 405 both structured and unstructured data. For example, the aggregated data 406 produced by aggregation module 405 may aggregate the structured data presented in FIG.5 with the unstructured data provided in Tables 1 and 2 and FIG.6. In addition, the aggregated data 406 may aggregate various forms of classes of unstructured data such as, for example, coordinate-based data (e.g., the data presented in Table 1), data sheet-based data (e.g., the data presented in Table 1), and composite data such as narrative-based data (e.g., the data 470 presented in FIG.6).

[0060] In some embodiments, the aggregated data 406 formed by aggregation module 405 from the separate classes of facility data 390-394 can all be associated or related to one another via association with one or more common references whereby qualitative analysis may be performed on the aggregated data 406 produced by aggregation module 405. Particularly, aggregation module 405 may utilize one or more different data processing techniques and / or structures to link together the separate classes of facility data 390-394when forming aggregated data 406. In some embodiments, aggregation module 405 may integrate different classes of data when forming aggregated data 406 into a data structure that links together (e.g., geometrically) the aggregated data 406. For instance, the data structure may store and maintain the relationships between the various entities (e.g., equipment of the operational facility) related to the aggregated data 406.

[0061] In some embodiments, the data structure comprises a knowledge graph. However, the configuration of the data structures implemented by dynamic digital replica 400 (replica 400 may implement one or more different data structures) may vary depending on the given application. For example, in some embodiments, the aggregated data 406 produced by aggregation module 405 may be maintained using a network of distributed relational databases, or distributed file systems such as HDFS (Hadoop Distributed File System), or some combination thereof.

[0062] Referring now to FIG. 7, an exemplary knowledge graph 480 associated with an operational facility and producible by dynamic digital replica 400 (e.g., by aggregation module 405) is shown. In this example, knowledge graph 480 includes one or more nodes 482 and one or more directional edges 484 that interconnect the nodes 482. Edges 484 are directional and thus extend from a beginning point to an end point to thereby represent the direction of a given edge 484 extending between a pair of nodes 482 interconnected by the edge.

[0063] The nodes 482 of knowledge graph 480 represent various entities of a given operational facility associated with or referenced by the data ingested by aggregation module 405. For example, in the example of FIG.7, nodes 482 of knowledge graph 480 represent personnel (e.g., nodes 482-1 and 482-2), structures (e.g., nodes 482-3 and 482- 5), equipment (e.g., nodes 482-6, 482-7, and 482-9), predefined actions or processes (e.g., nodes 482-4 and 482-8), and data (482-10) each associated with a given operational facility.

[0064] The edges 484 of knowledge graph 480 link the various nodes 482 together. Particularly, edges 484 indicate and represent the particular relationship linking together the given nodes 482 connected together by the edge 484. The nature of the relationship may vary widely. For example, the relationship represented by a given edge 484 may specify an identity (e.g., a first node “is a” second node as indicated by edges 484-1 and 484-5 of knowledge graph 480). An edge 484 may indicate that a first node 482 does,performs, or uses a given action or thing specified by a second node 482 (e.g., nodes 484- 3 and 484-4). An edge 484 may indicate that a first node 482 includes a second node 482 (e.g., edges 484-6, 484-7, 484-8, 484-9, and 484-11). An edge 484 may indicate that a first node 482 provides, generates, or creates a second node 482 (e.g., edge 484-10). Further, an edge 484 may indicate a working relationship between a first node 482 and a second node 482 (e.g., edge 484-2).

[0065] Unstructured data ingested by aggregation module 405 that is indirectly related or comprising many independent fields (e.g., the unstructured data presented in Table 1) can be related by knowledge graph 480 by creating new edges 484 defining relationships between nodes 482 (existing or newly added) of the knowledge graph 480. As an example, coordinates obtained from coordinate-based or composite data can be represented through the creation of new nodes 482 linked by edges 484 discernable from the from coordinate-based or composite data.

[0066] In some embodiments, unstructured data ingested by aggregation module 405 and containing text-intensive or narrative-based sources (e.g., the unstructured data presented in Table 2 and / or the data presented in FIG. 7) may be initially processed using Natural Language Processing (NLP) algorithms such as Word Sense Induction (WSI) and / or Word Sense Disambiguation (WSD). WSI is used to determine a sense inventory, which puts words into context. WSD compares the sense inventory with one from the knowledge graph to determine how the data is to be related. For example, a text corpus and a knowledge graph may be associated in a unified manner where, while both contain large amounts of data, the knowledge in the text corpus is usually implicit and unstructured, while the knowledge in knowledge graphs is explicit and structured.

[0067] Referring now to FIG. 8, a block diagram of an aggregation process 500 implementable by an aggregation module of a dynamic digital replica (e.g., aggregation module 405 of dynamic digital replica 400) is shown according to some embodiments. Generally, aggregation process 500 comprises several sequential stages including data ingestion stage 501, data processing stage 510, and the production 520 (e.g., creation, maintenance, updating) of one or more data structures 522 whereby qualitative analysis may be performed on the information presented and organized by the data structure 522. In this example, data ingestion stage 501 of aggregation process 500 includes ingesting several classes of data including both structured and unstructured data. Particularly, in thisexemplary embodiment, data ingestion stage 501 includes ingesting structured time-series data 502, unstructured coordinate-based data 503, unstructured data sheet-based data 504, and unstructured narrative-based data 505. However, the number and particular classes of data ingested by aggregation process 500 may vary in other embodiments.

[0068] Data processing stage 510 of aggregation process 500 includes the processing and aggregation of the ingested data 502-505 to form or produce aggregate data 511. In this exemplary embodiment, data processing stage 510 includes temporal-to-equipment association 512, coordinate-to-equipment association 513, and natural language processing 514. Each of processes 512, 513, and 514 may be performed by the aggregation module 405 of the dynamic digital replica 400 shown in FIG.4. The temporal- to-equipment association associates or links the time-series data 502 (e.g., flowrate data, pressure data, temperature data) with one or more related pieces of equipment (e.g., a pressure vessel to which the flowrate, pressure, and / or temperature data pertain) of an operational facility. In addition, the coordinate-to-equipment association 513 similarly associates or links the coordinate-based data 503 with one or more related pieces of equipment of an operational facility. For instance, coordinates obtained from coordinate- based data 503 may be matched with a particular piece of equipment via coordinate-to- equipment association 513 to thereby identify the coordinates in physical space of the given piece of equipment. Further, natural language processing 514 processes (e.g., via WSI, WSD, and / or other NLP techniques) text contained in the narrative-based data 505 whereby the processed text may be searched for or operated on.

[0069] The aggregated data 511 produced by data processing stage 510 of aggregation process 500 is organized in accordance with data structure 522 such that, in this example, the various entities extracted from data 502-505 and their varying relationships may be presented in a graphical format. As will be discussed further herein, the organization provided to aggregated data 511 by data structure 522 assists in facilitating the performance of qualitative analysis on the organized aggregated data 511.

[0070] Referring again to FIG. 4, the aggregated data 406 produced by aggregation module 405 is provided to a qualitative analysis module 410 of dynamic digital replica 400. Qualitative analysis module 410 is generally configured to automatically (e.g., without human intervention) performs qualitative analysis on the aggregated data 406 and thereby produce one or more extracted features 412 comprising features or entities extracted fromthe separate classes of facility data 390-394. In some embodiments, each of the extracted features 412 is based on a plurality of the separate classes of facility data 390-394.

[0071] With conventional digital replicas of operational facilities, analysts will typically manually review the different aggregated data sources, a laborious exercise having limited efficacy. Particularly, aside from the complexity of correlating the various data sources, many issues which occur with the given operational facility (particularly long-term issues) are not visible from the data themselves and thus may be overlooked by such manual analysis. For example, data may indicate that fluid temperature in a particular fluid conduit of the operational facility is within operational limits but, however, the sustained operation of the fluid conduit at one extreme of these limits (e.g., high or low temperature), combined with climate and / or other factors, may result in unexpected material fatigue of the fluid conduit. Thus, qualitative analysis, such as that conducted by qualitative analysis module 410, facilitates identifying potential issues or other scenarios by extracting features (e.g., extracted features 412) from the underlying data.

[0072] Referring now to FIG. 9, a block diagram of a qualitative analysis process 530 implementable by a qualitative analysis module of a dynamic digital replica (e.g., qualitative analysis module 415 of dynamic digital replica 400 shown in FIG.4) is shown according to some embodiments. In some embodiments, the functions of qualitative analysis process 530 may be implemented by a plurality of modules including, for example, both an aggregation module (e.g., aggregation module 405) and a qualitative analysis module (e.g., qualitative analysis module 410).

[0073] Generally, similar to aggregation process 500, qualitative analysis process 530 comprises several sequential stages including data ingestion stage 531, data processing stage 540, and data contextualization process 548. In this example, data ingestion stage 531 of qualitative analysis process 530 includes ingesting several classes of aggregated data including time-series data 532, image data 533 (e.g., images, composite data including one or more images), and textual data 534 (e.g., text documents, composite data including textual data). However, the number and particular classes of data ingested by qualitative analysis process 530 may vary in other embodiments. Aggregated data 532, 533, and 534 may be obtained from knowledge graph 522 in this exemplary embodiment.

[0074] The data ingested at data ingestion stage 531 is processed via data processing stage 540 of qualitative analysis process 400 to thereby provide extracted features 545comprising elements or features extracted from the data 532-534. In this exemplary embodiment, data processing stage 540 includes a discrete Fourier transform (DFT) 541 configured to output one or more extracted features 545 from the time-series data 532. Particularly, a sliding window of width n may be applied to the time-series data 532 and the resulting n values for the sliding window may be transformed using DFT to obtain an extracted feature in the form of a feature vector. The feature vector is stored and the sliding window is translated across the time-series data 532 (e.g., one step forward in time) and this process is repeated to determine an additional feature vector that is also stored. The resulting plurality of feature vectors determined from the application of a plurality of temporally translated sliding windows may be used to identify various issues or scenarios of the operational facility and reflected in the time-series data 532. These scenarios may correspond to normal operation of a given piece of equipment of the operational facility or potential failure situations. In addition, DFT 541 may be continually applied to time-series data 532 as it is generated (e.g., in real-time or near real-time) and additional feature vectors may be determined and compared against previously determined feature vectors.

[0075] Referring briefly to FIG.10, an exemplary DFT process 525 is shown. Particularly, FIG. 10 illustrates a graph 526 depicting time series-data 527 (e.g., an operational parameter like temperature, pressure, etc., depicted over time). Time-series data 527 is formed from a plurality of different decomposed signals comprising sinusoids of varying amplitudes, frequencies, and phase angles associated with their specific contribution to the time-series data 527. A desired feature set 528 e.g.( [a11,…a1n]) may be extracted from the decomposed signals (or alternatively on the time-series data 527 itself) present in each of a plurality of separate and distinct time windows 529. The feature extraction can be on the original time-series value or, more likely, on the decomposed signals.

[0076] Returning to FIG. 9, feature vectors extracted through the application of DFT 541 may facilitate the identification of various types of defects in a mechanical assembly or components (e.g., a bearing) via vibration data collected by one or more vibration sensors coupled to the mechanical component. For instance, various types of roller bearing defects, such as defects in the rolling elements thereof, bearing race defects, etc., exhibit varying frequency distributions. However, analysis of the vibration data itself is difficult given that the energy of the vibration data caused by relatively weak or minor defects is often generally low and thus may be masked by noise present in the vibration data. Inorder to address this issue, features may be extracted from the vibration data in order to more easily and effectively identify even minor defects in the roller bearing (or other mechanical component). For example, the vibration data may be subjected to DFT 541 whereby a plurality of feature vectors may be obtained and analyzed in order to identify possible defects including minor defects that were previously masked within the vibration data.

[0077] In this exemplary embodiment, data processing state additionally includes binary classification 542. Particularly, binary classification 542 may be applied to the image data 533 in order to provide one or more extracted features 545 from the image data 533. The extracted features 545 in this instance may comprise classifications applied to the image data 533 via binary classification 542. As but one example, binary classification 542 may be applied to image data 533 pertaining to a fluid containing component (e.g., a fluid conduit, a pressure vessel) where the binary classification 542 applies either a “corroded” classification to the image data 542 or “not corroded” classification to the image data 542 to thereby determine whether corrosion has occurred to the given fluid containing component.

[0078] DFT 541 and / or binary classification 542 may employ or comprise machine learning (ML) algorithms including. In some embodiments, the ML algorithms of DFT 541 and binary classification 542 comprise supervised learning algorithms where training datasets are used to define scenarios or issues of the operational facility that future extracted features may be compared against. For example, in the case of identifying roller bearing defects, historical vibration data associated with documented failures may be compared with extracted features 545. In the example of corrosion identification, large numbers (e.g., thousands) of images of corroded and non-corroded components may be used to train the binary classifier employed in binary classification 542.

[0079] Further, in this exemplary embodiment, data processing stage 540 additionally includes NLP 543. Particularly, NLP 543 may be applied to the textual data 534 in order to provide one or more extracted features 545 from the textual data 534. NLP 543, like DFT 541 and binary classification 542, may employ or comprise one or more ML algorithms. In some embodiments, NLP 543 employs or comprises WSI and / or WSD.

[0080] The extracted features 545 obtained by data processing stage 540 are provided to data contextualization process 548 which, by consulting the information contained in datastructure 520, may transform the extracted features 545 into meaningful information that can be used to make operational decisions.

[0081] Referring again to FIG.4, in this exemplary embodiment, dynamic digital replica 400 includes a contextualization module 415 that receives the extracted features 412 produced by qualitative analysis module 410. Particularly, contextualization module 415 is generally configurated to generate contextualized data 416 by contextualizing the aggregated data 406 and extracted features 412 using contextual data 417 (which may be unstructured data) that is separate from (e.g., separately sourced from) the aggregated data 406. The contextualized data 416 produced by contextualization module 415 may facilitate the determination of operational decisions pertaining to a given operational facility.

[0082] In some embodiments, the separate contextual data 417 used by contextualization module 415 to contextualize the aggregated data 406 and extracted features 412 is operational experience data that has occurred (e.g., captured in document form by experienced personnel of the operational facility) gradually over long periods of time. As but one example, operational experience data may relate to historical lessons regarding the performance of particular pieces of equipment when operating within a given operational envelope in a given environment and the interactions that may occur between the environment and the piece of equipment (e.g., one set of operational parameters may be safe in a first environment but hazardous in a second environment whereby adjustments to the operational parameters will need to be made). Consequence Assessment Health & Safety Environmental Business Financial Business Non-Financial k nd ndTABLE 3

[0083] Table 3 illustrates exemplary contextual data in the form of operational experience data usable by contextualization module 415. Particularly, the operational experience data presented in Table 3 is in the form of a consequence assessment table. Given that the operational experience data indicated in Table 3 is in document form, it can be automatically ingested or used by the contextualization module 415. The example presented in Table 3 indicates the relationship between defects observed in walls of fluid containing components (e.g., fluid conduits, pressure vessels) and the consequence of the defect in terms of health and safety, environmental impact, and business impact. The body of knowledge in this area, based on real events over many years, has been distilled in Table 3 by experienced personnel of the operational facility into a rating level that can be used to categorize defects so that they can be addressed in order of potential consequence. In this manner, Table 3 recapitulates operational experience data collected over a substantial period of time.

[0084] It may be noted in Table 3 that the operational experience data captured therein indicates that the nature of the consequence is not based only on the size or scope of the defect, and instead is contingent on a broad array of additional information. For example, the nature of the consequence may be contingent on, for example: (1) the location of the defect (trackable via 3D models and 3D general assembly data); (2) equipment design parameters (obtainable from equipment data sheets); (3) historical usage of the equipment (obtainable from control system historical trend data reports); (4) maintenance history of the equipment (obtainable via maintenance reports); (5) historical weather conditions and future weather forecasts (both short-term and long-term); and (6) results of what-if- scenarios when taking the given equipment out of service for repair or replacement. By consulting these different classes of contextual data, meaningful and accurate recommendations to personnel of the operational facility may be made automatically such that the personnel need not attempt to deduce the same insights manually. In addition, the contextual data 417 used by contextualization module 415 comprises textual data that may be subjected to NLP to permit the textual data to be aligned with the aggregated data 406 and extracted features 412 processed by the contextualization module 415.

[0085] The classes or types of contextual data 417 used by contextualization module 415 may depend, beyond the configuration of the operational facility and the availability of contextual data pertaining to the facility, on the specific objective of the dynamic digital twin400. Therefore, in some embodiments, contextualization module 415 employs Principal Component Analysis (PCA) to reduce the dimensionality of the ingested dataset (e.g., aggregated data 406 and / or extracted features 412) to that which is required to yield an optimal conclusion in view of the specific objective of dynamic digital replica 400.

[0086] Briefly, contextualization module 415 may implement PCA (following qualitative analysis performed by module 410) by computing a covariance matrix of the analyzed aggregated data 412, where the covariance matrix describes the relationships between pairs of features of the analyzed aggregated data 406. In addition, eigenvectors and eigenvalues of the covariance matrix may be calculated which represent the directions and magnitudes of the principal components of the covariance matrix (the eigenvectors corresponding to the principal components and the eigenvalues indicating their importance). With the eigenvectors and eigenvalues determined, the eigenvalues may be sorted in descending order to identify their most significant principal components which help explain the most variance in the underlying dataset (e.g., aggregated data 406). Subsequently, the number of principal components to retain may be determined which may be based on a cumulative explained variance threshold, domain knowledge, or other bases. Finally, the selected eigenvectors may be used to transform the analyzed aggregated data 412 into a lower-dimensional space which may involve matrix multiplication to obtain the principal components’ values for each data point. The resulting lower-dimensional, contextualized data 416 may comprise linear combinations of the aggregated data 406 and may reveal underlying patterns and relationships within the aggregated data 406. Thus, in at least some embodiments, the contextualized data 416 produced by contextualization module 415 may have a dimensionality that is less than the dimensionality of the aggregated data 406 ingested by the contextualization module 415 where the reduction in dimensionality may be based on the specific objective of the dynamic digital replica 400.

[0087] In this exemplary embodiment, dynamic digital twin 400 includes a self-learning system 402 that receives the contextualized data 416 produced by contextualization module 415. Generally, the overall or generalized objective of dynamic digital twin 400 is to provide useful recommendations and insights to an operator of the operational facility corresponding to the replica 400. Generally, there may be several distinct decision variables such as, for example, specific objectives, costs, and time. Recommendationsprovided by dynamic digital twin 400 correspond to an optimal combination of decision variables based on certain constraints. In some embodiments, Linear Programming (LP) is employed by self-learning system 400 to assist in generating recommendations corresponding to the optimal combination of decision variables.

[0088] In some embodiments, dynamic digital twin 400 is initially trained using known datasets and knowledge (e.g., data 390-394). Generally, the self-learning system 402 may use this training to produce recommendations which may be fed-back to the underlying datasets to improve the underlying knowledge base. In this manner, self-learning system 402 permits the dynamic digital twin 400 to continuously optimize itself based on, for example, the contextualized data 416 (e.g., including the dimensionality defined by PCA).

[0089] In at least some situations, self-learning system 402 determines, based on contextualized data 416, whether the operational facility is in a known state 418 or an unknown state 419. If in the known state 418, a recommendation module 420 of self- learning system 402 may produce a recommendation 422 (e.g., for addressing a specific issue or scenario) based on the contextualized data 416 provided by contextualization module 415. The recommendation 422 may pertain to an action to be performed on at least one of plurality of equipment of the operational facility to which the dynamic digital replica 400 corresponds. The recommendation 422 may be provided to a user of the dynamic digital replica 400 via a user interface. In some embodiments, the recommendation 422 may be provided to an operator or owner of the operational facility corresponding to the dynamic digital replica 400.

[0090] In some embodiments, the recommendation 422 provided by recommendation module 420 corresponds to an optimal combination of decision variables based on certain constraints. The optimal combination of decision variables may be based at least partly on a predefined specific objective of the dynamic digital replica 400. In some embodiments, the recommendation 422 may be generated by applying LP to the contextualized data 416 provided to recommendation module 420. Additionally, in this exemplary embodiment, recommendation 422 is provided to a user 401 of the dynamic digital replica 400. The user 401 may perform a recommended action 403 (e.g., to or upon the operational facility) based on the recommendation 422 generated by the recommendation module 420.

[0091] In some embodiments, the user may provide feedback data 423 to the aggregation module 405 whereby the aggregation module 405 may aggregate the feedback data 423with the different classes of data 390-394 supplied to the aggregation module 405. The feedback data 423 may be provided by the user 401 in response to or based on the recommendation 422. By aggregating feedback data 423 with the other classes of data 390-394 additional insights (e.g., with regard to the response of the operational facility in response to performing an action to the facility based on the given recommendation 422) may be gleaned by dynamic digital replica 400 to improve the quality of future recommendations 422 provided by recommendation module 420.

[0092] In some instances, a scenario or issue encountered by dynamic digital replica 400 may unknown such that the operational facility is an unknown state 419 such that recommendation module 420 is unable to formulate a recommendation 422 based on the contextualized data 416. For example, in some instances dynamic digital replica 400 may be unable to ascertain why a given scenario or issue has occurred and thus cannot formulate a recommendation 422. In some embodiments, when an unknown state 419 of the operational facility is identified by the dynamic digital replica 400, replica 400 searches for additional data to assist the replica 400 in determining the cause behind the given scenario.

[0093] In this exemplary embodiment, the self-learning system 402 of dynamic digital replica 400 includes a search module 425 configured, when replica 400 encounters an unknown state 419, to search for additional data when replica 400 that may permit the replica 400 to make a recommendation 422 addressing a previously undecipherable scenario thereby placing the operational facility into the known state 418. In some embodiments, additional data uncovered by search module 425 may be used to update or refine the PCA performed by contextualization module 415 resulting in the updating or refining of contextualized data 416 (e.g., via the obtainment of a new set of dimensions from the refined PCA). To state in other words, the additional data uncovered by search module 425 may be used by contextualization module 415 to unearth new correlations between data captured in aggregated data 406.

[0094] As an example, a scenario may arise in an operational facility whereby equipment of the operational facility corrodes faster than predicted and the dynamic digital replica 400 associated with the given operational facility is unable to initially determine the reason why. In such a scenario, search module 425 may search existing datasets (e.g., aggregated data 406) or new datasets (e.g., information available in other systems or via the Internet)to identify new correlations between certain data and the rate of equipment corrosion and, by updating its dimensionality through PCA conducted via the contextualization module 415, automatically include these newly discovered correlations in future analyses performed by the dynamic digital replica 400.

[0095] In this exemplary embodiment, self-learning system 402 further includes a query module 430 that may query a user of the dynamic digital replica 400 (e.g., a subject matter expert (SME) of the given operational facility) for input should the operational facility remain in the unknown state 419 following activation of the search module 425. In other words, should the dynamic digital replica 400 remain unable to formulate a recommendation 422 pertaining to a scenario or issue encountered by the operational facility such that the operational facility. In this manner, dynamic digital replica 400 may capture further domain knowledge that is not fully documented anywhere (e.g., within aggregated data 406) and instead may reside only within the head of the given user or experienced personnel of the operational facility. Aggregated data 406 and contextualized data 416 may each be automatically updated or refined by aggregation module 405 and contextualization module 415, respectively, once this additional knowledge is captured from the user (e.g., through an appropriate user interface) by dynamic digital replica 400 (e.g., new data sources incorporated, new rules captured) to account for these changes. The updated aggregated data 406 and contextualized data 416 may be used by self-learning system 402 to discover the reasons behind the given scenario thereby permitting the recommendation module 420 to formulate a recommendation 422 with the operational facility being in the known state 418.

[0096] To further illustrate the functionalities of embodiments of dynamic digital replicas disclosed herein, an example is presented focusing on asset integrity management of an operational facility in the form of an offshore oil and gas production facility. Asset integrity management encompasses the design, operation, and maintenance of an asset (e.g., one or more pieces of equipment of an operational facility) to preserve its integrity at an acceptable level of risk throughout its operational life. Protection of health, safety, and the environment are important components of the processes and procedures used to monitor the conditions of offshore production facilities as well as other operational facilities (e.g., offshore surface and subsea facilities and structures).

[0097] The assessment of an operational facility’s asset integrity status is often performed using a method commonly referred to as barrier assessment that utilizes a model called bowtie or Swiss cheese. In either model, multiple controls or barriers are identified to protect an operational facility against a particular incident (e.g., loss of primary containment of a piece of equipment such as a pressure vessel) causing a major accident (e.g., a fire or explosion). Some controls are preventative in that the preventive controls are configured to prevent an incident from occurring (e.g., a pressure release valve of a pressure vessel configured to activate in the event of a potential over-pressurization of the vessel). Alternatively, a given control may be a mitigating control configured to minimize or reduce negative consequences of an incident (e.g., fire detectors configured to quickly detect the presence of a fire).

[0098] Referring now to FIG.11, a block diagram of an exemplary bowtie model 550 of an operational facility in the form of an offshore production facility is shown according to some embodiments. Bowtie model 550 presents a set of potential incidents or scenarios pertaining to a given asset (e.g., a pressure vessel of the offshore production facility) – an overpressure 551, an overfill 560, and external corrosion 570. In addition, bowtie model 550 identifies one or more barriers 552, 562, and 572 specifically associated with preventing the potential scenarios 551, 560, and 570, respectively causing the event 580. These barriers 552, 562, and 572 include, for example, a process control system 552-1 (e.g., to identify a controller that reads pressure and automatically opens a valve to control this value), an operator response 562-2, and fabric maintenance 572-3.

[0099] Further, bowtie model 550 includes one or more barriers that mitigate the consequences of the event 580 if it occurs. In this example, event 580 comprises loss of primary containment (LOPC) associated with flammable material present in the vessel. In this exemplary embodiment, bowtie model 550 includes mitigation barriers such as emergency system shutdown 580-1 and active fire protection 580-4. Bowtie model 550 further includes a consequence 590 that may result should each of the barriers 580-1 to 580-4 fail to mitigate the event 580. In this exemplary embodiment, the consequence 590 resulting from failure to mitigate event 580 (e.g., failure to contain the LOPC) comprises fire or an explosion of the vessel.

[0100] The overall aim of asset integrity is to ensure that all barriers (e.g., barriers 552, 562, 572, and 580) are always healthy and fully operational. Each of the barriers (e.g.,barriers 552, 562, 572, and 580) of bowtie model 550 has an associate dataset that may be analyzed to determine the health of the given barrier. For example, in the case of process control system 552-1, the associated dataset may include maintenance records for the associated controller and valve to ensure that preventative and corrective maintenance has been performed within specified timescales. In addition, the associated dataset may include historical real-time data to determine if any process excursions of the vessel have occurred in the past.

[0101] Computing an entire bowtie model may be challenging and resource intensive because the data required is in various formats and captured in a plurality of separate systems. In addition, some information may not be readily available in conventional systems. For example, hazardous location certified equipment in proximity to a particular vessel may not be captured in a searchable system. However, embodiments of dynamic digital replicas (e.g., dynamic digital replica 400) disclosed herein may incorporate 3D representations of the operational facility and links to equipment attributes, thereby making it possible to automatically identify devices of a particular classification that are within a certain distance of the vessel.

[0102] Corrosion is a major concern in asset integrity management, especially for those operational facilities located in harsh environments such as offshore facilities. To manage the risk, asset owners often have inspection programs using various methods such as close visual inspection and general visual inspection where inspectors take photographs and record other observations (e.g., x-ray and ultrasonic measurements).

[0103] Due to cost and limitations in space offshore (commonly referred to as Personnel on Board, or POB), asset owners of operational facilities in the form of offshore production facilities often create an inspection schedule that appropriately manages the corrosion risk in the most cost-effective manner. This typically means that the given operational facility is under permanent inspection such that once the entire facility has been inspected it is time to begin a new inspection.

[0104] This conventional practice may be acceptable for managing asset integrity risk in some applications, but it has many limitations. For example, such conventional practices are difficult to scale up and inspect at a faster rate, or to dynamically adjust the inspection schedule to meet changing asset integrity, POB, or other constraints. In addition, inspection results are dependent on the particular individuals involved. Particularly, asidefrom some quantitative measurements, many aspects of inspection involve the use of individual judgement by the given inspector contingent on the specific experience and competence of the inspector. Further, it is possible that asset integrity failures can occur due to corrosion in areas that are not due for inspection. Such a failure can be catastrophic, and at the very least may be an extremely costly incident to address.

[0105] Embodiments of dynamic digital replicas (e.g., dynamic digital replica 400) disclosed herein are configured to address these limitations of conventional practices. Particularly, embodiments of dynamic digital replicas disclosed herein automate aggregation and contextualization of a plurality of separate data sources making it possible to change criteria and obtain a new results in real-time or near-real time. In addition, embodiments of dynamic digital replicas disclosed herein employ computer implemented analysis techniques (e.g., PCA employed by contextualization module 415) eliminates individual judgement that contributes to inconsistent results. In addition, embodiments of dynamic digital replicas disclosed herein employ automated techniques that are not dependent on the number of available personnel (e.g., the number of available inspectors), and so the rate of inspection can be increased to reduce the likelihood of an asset integrity failure between inspections. Further, embodiments of dynamic digital replicas disclosed herein may include self-learning systems (e.g., self-learning system 402) that allow continuous improvement in asset integrity performance.

[0106] Referring to FIG. 12, a block diagram of another dynamic digital replica 600 is shown according to some embodiments. Dynamic digital replica 400 is used for asset integrity management for an operational facility. In this exemplary embodiment, dynamic digital replica 600 includes separate classes of facility data 601-605 including 3D models 601, equipment attributes 602, process diagrams 603, construction drawings 604, and surface scans in the form of laser scans and photogrammetry 605.

[0107] A basic digital replica 610 associated with the operational facility may be constructed from the separate classes of facility data 601-605. In this exemplary embodiment, dynamic digital replica 600 includes a corrosion analytics module 615 for performing corrosion analysis using data provided by the basic digital replica 610. In some embodiments, corrosion analytics module 615 uses a binary classification algorithm to generate a list of areas of concern identified by geospatial coordinates. Additionally, corrosion analytics module 615 may determine the severity of any extant corrosion basedon 3D measurements of the laser scans and photogrammetry 605. Further, in some embodiments, corrosion analytics module 615 may identify equipment of the operational facility affected by corrosion, the criticality of that equipment (e.g., based on the function the equipment performs), and employ this information with the determined severity of the corrosion to generate an overall ranking of the areas of potential corrosion (e.g., high, medium, or low concern).

[0108] In this exemplary embodiment, dynamic digital twin 600 includes an equipment degradation forecast module 620 that combines results from the corrosion analysis performed by corrosion analytics module 615 with additional contextual data to develop an equipment degradation forecast 621. In some embodiments, the equipment degradation forecast module 620 uses contextual data including real-time process data 622, weather conditions 623, manufacturer data 624, and inspection records 625 in formulating the equipment degradation forecast 621. In this exemplary embodiment, equipment degradation forecast 621 informs the rate at which corrosion is occurring and the likely rate it will continue to occur, based on usage. This data may be combined using algorithms described elsewhere herein including, for example, DFT and NLP.

[0109] In this exemplary embodiment, dynamic digital twin 600 additionally includes a barrier status module 630 that determines the barrier status 631 for one or more (or each) barriers of the operational facility. In some embodiments, barrier status module 630 employs layer of protection analysis (LOPA) 635 in determining the barrier status 631 of the barriers of the operational facility. Generally, LOPA 635 is a structured and systematic technique used to assess and analyze the effectiveness of layers of protection designed to prevent or mitigate the consequences of an incident. Particularly, LOPA 635 may involve identifying and evaluating various layers of protection or barriers such as safety systems, alarms, procedures, and inherent process features, to ensure that the overall risk of a potential incident is reduced to an acceptable level.

[0110] From the results of the activation of barrier status module 630, a list of required activities (e.g., repairing a given valve, replacing a given pump, inspecting a given fluid conduit) forming a maintenance schedule 640 may be derived. In addition, LP may be applied to the maintenance schedule 640, the barrier status 631, and an existing POB schedule 645 to determine the optimal execution of these activities to manage asset integrity risk in the most cost-effective manner. In other words, feedback data 646 may berelayed from the POB schedule 645 to the maintenance schedule 640. In addition, feedback data 641 may be relayed from the maintenance schedule 640 to the barrier status module 630.

[0111] Referring now to FIG. 13, an embodiment of a computer system 700 is shown suitable for implementing one or more components disclosed herein. As an example, computer system 700 may be used to execute various embodiments of material testing systems (e.g., material testing system 10 shown in FIG. 1) disclosed herein. As an example, the computer system 700 may comprise an embodiment of the system controller shown in FIG.1.

[0112] The computer system 700 of FIG. 13 generally includes a processor 702 (which may be referred to as a central processor unit or CPU) that is in communication with memory devices including secondary storage 704, read only memory (ROM) 706, random access memory (RAM) 708, input / output (I / O) devices 710, and network connectivity devices 712. The processor 702 may be implemented as one or more CPU chips. It is understood that by programming and / or loading executable instructions onto the computer system 700, at least one of the CPU 702, the RAM 708, and the ROM 706 are changed, transforming the computer system 700 in part into a particular machine or apparatus having the novel functionality taught by the present disclosure.

[0113] Additionally, after the system 700 is turned on or booted, the CPU 702 may execute a computer program or application. For example, the CPU 702 may execute software or firmware stored in the ROM 706 or stored in the RAM 708. In some cases, on boot and / or when the application is initiated, the CPU 702 may copy the application or portions of the application from the secondary storage 704 to the RAM 708 or to memory space within the CPU 702 itself, and the CPU 702 may then execute instructions that the application is comprised of. In some cases, the CPU 702 may copy the application or portions of the application from memory accessed via the network connectivity devices 712 or via the I / O devices 710 to the RAM 708 or to memory space within the CPU 702, and the CPU 702 may then execute instructions that the application is comprised of. During execution, an application may load instructions into the CPU 702, for example load some of the instructions of the application into a cache of the CPU 702. In some contexts, an application that is executed may be said to configure the CPU 702 to do something, e.g., to configure the CPU 702 to perform the function or functions promoted by the subjectapplication. When the CPU 702 is configured in this way by the application, the CPU 702 becomes a specific purpose computer or a specific purpose machine.

[0114] Secondary storage 704 may be used to store programs which are loaded into RAM 708 when such programs are selected for execution. The ROM 706 is used to store instructions and perhaps data which are read during program execution. ROM 706 is a non-volatile memory device which typically has a small memory capacity relative to the larger memory capacity of secondary storage 704. The secondary storage 704, the RAM 708, and / or the ROM 706 may be referred to in some contexts as computer readable storage media and / or non-transitory computer readable media. I / O devices 710 may include printers, video monitors, liquid crystal displays (LCDs), touch screen displays, keyboards, keypads, switches, dials, mice, track balls, voice recognizers, card readers, paper tape readers, or other well-known input devices.

[0115] The network connectivity devices 712 may take the form of modems, modem banks, Ethernet cards, universal serial bus (USB) interface cards, wireless local area network (WLAN) cards, radio transceiver cards, and / or other well-known network devices. The network connectivity devices 712 may provide wired communication links and / or wireless communication links. These network connectivity devices 712 may enable the processor 702 to communicate with the Internet or one or more intranets. With such a network connection, it is contemplated that the processor 702 might receive information from the network, or might output information to the network. Such information, which may include data or instructions to be executed using processor 702 for example, may be received from and outputted to the network, for example, in the form of a computer data baseband signal or signal embodied in a carrier wave.

[0116] The processor 702 executes instructions, codes, computer programs, scripts which it accesses from hard disk, floppy disk, optical disk, flash drive, ROM 706, RAM 708, or the network connectivity devices 712. While only one processor 702 is shown, multiple processors may be present. Thus, while instructions may be discussed as executed by a processor, the instructions may be executed simultaneously, serially, or otherwise executed by one or multiple processors. Instructions, codes, computer programs, scripts, and / or data that may be accessed from the secondary storage 704, for example, hard drives, floppy disks, optical disks, and / or other device, the ROM 706, and / or the RAM 708may be referred to in some contexts as non-transitory instructions and / or non-transitory information.

[0117] In an embodiment, the computer system 700 may comprise two or more computers in communication with each other that collaborate to perform a task. For example, but not by way of limitation, an application may be partitioned in such a way as to permit concurrent and / or parallel processing of the instructions of the application. Alternatively, the data processed by the application may be partitioned in such a way as to permit concurrent and / or parallel processing of different portions of a data set by the two or more computers. In an embodiment, the functionality disclosed above may be provided by executing the application and / or applications in a cloud computing environment. Cloud computing may comprise providing computing services via a network connection using dynamically scalable computing resources.

[0118] Referring now to FIG. 14, a flowchart of a computer-implemented method 750 for generating a dynamic digital replica of an operational facility is shown according to some embodiments. Initially, at block 752 method 750 includes aggregating (e.g., by aggregation module 405 shown in FIG.4) separate classes of facility data (e.g., facility data 391-394 shown in FIG. 4) each pertaining to one or more of a plurality of equipment of the operational facility to thereby produce aggregated data (e.g., aggregated data 406 shown in FIG. 4). At block 754, method 750 comprises performing qualitative analysis (e.g., by the qualitative analysis module 410 shown in FIG.4) on the aggregated data to extract one or more features (e.g., extracted features 412 shown in FIG. 4) associated with one or more of the plurality of equipment, wherein each of the extracted features is based on a plurality of the separate classes of facility data.

[0119] At block 756, method 750 comprises contextualizing (e.g., by the contextualization module 415 shown in FIG. 4) the aggregated data and the extracted features using unstructured data (e.g., contextual data 417 shown in FIG. 4) that is separate from the aggregated data to generate contextualized data (e.g., contextualized data 416 shown in FIG.4). At block 758, method 750 comprises generating a dynamic digital replica of the operational facility based on the aggregated data, the extracted features, and the contextualized data. At block 760, method 750 comprises providing, using the dynamic digital replica, a recommendation (e.g., recommendation 422 shown in FIG. 4) to a userpertaining to an action to be performed on at least one of the plurality of equipment of the operational facility.

[0120] While embodiments of the disclosure have been shown and described, modifications thereof can be made by one skilled in the art without departing from the scope or teachings herein. The embodiments described herein are exemplary only and are not limiting. Many variations and modifications of the systems, apparatus, and processes described herein are possible and are within the scope of the disclosure. For example, the relative dimensions of various parts, the materials from which the various parts are made, and other parameters can be varied. Accordingly, the scope of protection is not limited to the embodiments described herein, but is only limited by the claims that follow, the scope of which shall include all equivalents of the subject matter of the claims. Unless expressly stated otherwise, the steps in a method claim may be performed in any order. The recitation of identifiers such as (a), (b), (c) or (1), (2), (3) before steps in a method claim are not intended to and do not specify a particular order to the steps, but rather are used to simplify subsequent reference to such steps.

Claims

CLAIMS What is claimed is:

1. A computer-implemented method for generating a dynamic digital replica of an operational facility, the method comprising: (a) aggregating separate classes of facility data each pertaining to one or more of a plurality of equipment of the operational facility to thereby produce aggregated data; (b) performing qualitative analysis on the aggregated data to extract one or more features associated with one or more of the plurality of equipment, wherein each of the extracted features is based on a plurality of the separate classes of facility data; (c) contextualizing the aggregated data and the extracted features using unstructured data that is separate from the aggregated data to generate contextualized data; (d) generating a dynamic digital replica of the operational facility based on the aggregated data, the extracted features, and the contextualized data; and (e) providing, using the dynamic digital replica, a recommendation to a user pertaining to an action to be performed on at least one of the plurality of equipment of the operational facility.

2. The method of claim 1, further comprising: (f) determining, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e); and (g) providing, using the dynamic digital replica, the recommendation to the user at (e) in response to determining at (f) that the operational facility is in the known state.

3. The method of claim 1, further comprising: (f) determining, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e); and (g) requesting, using the dynamic digital replica, a user for input pertaining to the operational facility following a determination at (f) that the operational facility is in the unknown state.

4. The method of claim 3, further comprising: (h) receiving as feedback data, using the dynamic digital replica, the requested input pertaining to the operational facility from the user to which input is requested at (g); (i) determining, using the dynamic digital replica, that the operational facility is in the known state in response to receiving the feedback data; and (j) providing the recommendation to the user at (e) in response to (i).

5. The method of claim 1, further comprising: (f) applying, using the dynamic digital replica, feedback data from a user to the aggregated data, wherein the feedback data is based on the recommendation made to the user at (e).

6. The method of claim 5, wherein (f) comprises using principal component analysis (PCA) to update a dimensionality of the aggregated data based on the feedback data.

7. The method of claim 1, further comprising: (f) determining, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e); and (g) searching, using the dynamic digital replica, one or more of the aggregated data, the extracted features, and the contextual data to generate additional contextualized data following a determination at (f) that the operational facility is in the unknown state.

8. The method of claim 7, further comprising: (h) determining, using the dynamic digital replica, that the operational facility is in the known state in response to producing the additional contextualized data at (g); and (i) providing the recommendation to the user at (e) in response to (h).

9. The method of claim 1, further comprising: (f) determining, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e);(g) searching, using the dynamic digital replica, one or more of the aggregated data, the extracted features, and the contextual data to generate additional contextualized data following a determination at (f) that the operational facility is in the unknown state; (h) determining, using the dynamic digital replica, that the operational facility remains in the unknown state following (g); and (i) requesting a from for a user input pertaining to the operational facility following the determination at (h) that the operational facility remains in the unknown state.

10. The method of claim 9, further comprising: (j) receiving, by the dynamic digital replica, as feedback data the requested input pertaining to the operational facility from the user from which input is requested at (i); (k) determining, using the dynamic digital replica, that the operational facility is in the known state in response to receiving the feedback data; and (l) providing the recommendation to the user at (e) in response to (k).

11. The method of claim 1, wherein (c) comprises perform a principal component analysis (PCA) on the aggregated data to reduce a dimensionality of the aggregated data to generate the contextualized data.

12. A computer-readable medium storing executable code which, when executed by a processor, causes the processor to: (a) aggregate separate classes of facility data each pertaining to one or more of a plurality of equipment of an operational facility to thereby produce aggregated data; (b) perform qualitative analysis on the aggregated data to extract one or more features associated with one or more of the plurality of equipment, wherein each of the extracted features is based on a plurality of the separate classes of facility data; (c) contextualize the aggregated data and the extracted features using unstructured data that is separate from the aggregated data to generate contextualized data;(d) generate a dynamic digital replica of the operational facility based on the aggregated data, the extracted features, and the contextualized data; and (e) provide, using the dynamic digital replica, a recommendation to a user pertaining to an action to be performed on at least one of the plurality of equipment of the operational facility.

13. The computer-readable medium of claim 12, when executed by the processor, causes the processor to: (f) determine, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e); and (g) provide, using the dynamic digital replica, the recommendation to the user at (e) in response to determining at (f) that the operational facility is in the known state.

14. The computer-readable medium of claim 12, when executed by the processor, causes the processor to: (f) determine, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e); and (g) request from a user input pertaining to the operational facility following a determination at (f) that the operational facility is in the unknown state.

15. The computer-readable medium of claim 14, when executed by the processor, causes the processor to: (h) receive, by the dynamic digital replace, as feedback data the requested input pertaining to the operational facility from the user from which input is requested at (g); (i) determine, using the dynamic digital replica, that the operational facility is in the known state in response to receiving the feedback data; and (j) provide the recommendation to the user at (e) in response to (i).

16. The computer-readable medium of claim 12, when executed by the processor, causes the processor to:(f) apply, using the dynamic digital replica, feedback data from a user to the aggregated data, wherein the feedback data is based on the recommendation made to the user at (e).

17. The computer-readable medium of claim 16, wherein (f) comprises using principal component analysis (PCA) to update a dimensionality of the aggregated data based on the feedback data.

18. The computer-readable medium of claim 12, when executed by the processor, causes the processor to: (f) determine, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e); and (g) search, using the dynamic digital replica, one or more of the aggregated data, the extracted features, and the contextual data to generate additional contextualized data following a determination at (f) that the operational facility is in the unknown state.

19. The computer-readable medium of claim 18, when executed by the processor, causes the processor to: (h) determine, using the dynamic digital replica, that the operational facility is in the known state in response to producing the additional contextualized data at (g); and (i) provide the recommendation to the user at (e) in response to (h).

20. The computer-readable medium of claim 12, when executed by the processor, causes the processor to: (f) determine, using the dynamic digital replica, whether the operational facility is in a known state or an unknown state prior to (e); (g) search, using the dynamic digital replica, one or more of the aggregated data, the extracted features, and the contextual data to generate additional contextualized data following a determination at (f) that the operational facility is in the unknown state; (h) determine, using the dynamic digital replica, that the operational facility remains in the unknown state following (g); and(i) request a from a user input pertaining to the operational facility following the determination at (h) that the operational facility remains in the unknown state.

Images