System and method for monitoring the health of physical assets using digital twins

Abstract

The present invention relates to a system and method for monitoring the health of physical assets using digital twins. The method comprises data processing on a local server or on the cloud, emitting automatic warnings when defined thresholds are exceeded and displaying information in real time by means of an information display panel. The system includes sensors for different variables, a data acquisition module, software based on predictive models and a graphical interface that displays the status of the asset and allows operating history to be viewed. The invention can be used in structural, thermal and hydraulic monitoring of critical assets, thereby improving operating continuity and facilitating decision-making based on reliable data by means of autonomous processing and automated warnings.

Description

SYSTEM AND METHOD FOR MONITORING THE HEALTH OF PHYSICAL ASSETS THROUGH DIGITAL TWINS DESCRIPTIVE MEMORY SCOPE

[0001] The present invention relates to a system and method for monitoring the health of physical assets through the use of digital twins. In particular, the invention enables the integrity of such assets to be monitored through a digital representation, simulating their behavior in real time. This system is designed to be applied to critical assets in various sectors, such as industrial infrastructure, pipeline networks, and thermal systems, among others. Through the simulation and analysis of data obtained from sensors, the system facilitates informed, real-time decision-making, thus ensuring the reliability and safety of the monitored assets.

[0002] For example, the technology can be implemented in structural monitoring, thermal monitoring, watershed monitoring, and piping network monitoring. BACKGROUND AND TECHNICAL PROBLEM

[0003] In the context of industrial processes that demand both high productivity and compliance with strict safety and reliability regulations, this technology offers an innovative solution for Physical Asset Health Monitoring. This technology enables the implementation of an advanced Asset Integrity Management (AIM) service, facilitating continuous asset monitoring by simulating their behavior through digital twins. Its main objective is to provide added value by enabling timely decision-making based on accurate, real-time data. The technology covers structural, hydrological, and thermal monitoring, as well as other critical aspects related to production, thus optimizing operational efficiency and ensuring asset safety.

[0004] Through modeling and the information received, the technology allows for monitoring structural, thermal, and other production-related behavior. DESCRIPTION OF PREFERRED MODALITIES

[0005] The technology includes several key components, as illustrated in Figure 1, which are described below. Asset Checkup and Diagnosis

[0006] The method and system of the invention comprise asset checking and diagnosis, which allows for a comprehensive assessment of assets based on their criticality. This criticality is determined by evaluating their importance within the production chain. The most relevant assets are prioritized for instrumentation, ensuring that resources are concentrated on the most operationally critical assets that require continuous monitoring. Survey and Modeling: Structural and theoretical analysis

[0007] The method and system of the invention comprise the collection of information necessary to construct the digital twin of the asset. This collection is carried out using plans, calculation memories, topographic equipment, a light detection and ranging (LIDAR) scanner, and other available information. The digital twin is used to simulate the asset's behavior for monitoring purposes. For example, the construction of a structural model to evaluate the behavior of an industrial building under the impact of extreme operating conditions.

[0008] An analysis is performed using the digital twin, developed using proprietary software that allows for determining the optimal sensor locations that provide the greatest amount of information for monitoring purposes. The sensors, which can be custom-designed, are designed to pose no risks during installation or to the operation of the asset or its production processes.

[0009] As a result of this stage and with the available information, structural models will be obtained and calibrated, allowing the structure's behavior to be evaluated based on the data received from the instrumentation. Sensor Installation: Parameter Definitions and Alerts

[0010] The method and system of the invention comprise the installation of sensors and communication equipment that allow for the assessment of impacts on the asset for monitoring purposes. These sensors can measure various variables, including environmental conditions, structural behavior, images, and videos. The location of the sensors is determined to maximize the collection of relevant data.

[0011] The system and method of the invention can simultaneously acquire data from multiple sensors, regardless of their nature, and process this information with microsecond precision, ensuring reliable and accurate monitoring.

[0012] The initial objective of the instrumentation is to calibrate the structural models created, and its specific objectives are to monitor variables indicative of the asset's condition. This allows, for example, to determine the structural integrity of the building / structure based on estimates of specific parameters, such as vibration frequencies, correlations between variables, monitoring of fatigue cycles, calculation of floor response spectra, among other characteristics. Instant Cloud Analytics: Constant Monitoring and Visualization

[0013] The technology includes asset monitoring. After the sensors are installed, continuous monitoring of the asset begins immediately, which can be performed in the cloud or on a local server. The collected information is analyzed and processed in compliance with applicable regulations and compared with the expected behavior of the asset simulated by the digital twin.

[0014] In this phase, the system uses specialized data acquisition software on the local server and a cloud-based processing engine, which is the analytics engine of the invention. This is done with the goal of performing real-time analysis and generating critical information for decision-making. Alerts: Protocols for possible failures

[0015] The method and system of the invention include an information display and alert generation system that organizes and presents the acquired and analyzed data through a dashboard. This dashboard graphically and intuitively displays, using a traffic light-type system, the condition of the monitored asset, as well as its ability to withstand operating conditions throughout its remaining useful life. The system is designed to display the asset's status in real time, allowing for rapid identification of potential problems.

[0016] The information deployment system is connected to an automated alert generation system, which sends real-time notifications to users through multiple channels, including email, SMS, WhatsApp, and Microsoft Teams. These alerts are automatically generated when data collected by the sensors exceeds predefined operational thresholds. This immediate response capability allows users to make informed decisions quickly to avoid potential operational or structural failures. Furthermore, A historical log is kept where all relevant information for the asset is published, events, alerts, reports, photos, where said log can be accessed by different user profiles (for example, administrator, operator, maintenance). Modalities of the system and method of the invention

[0017] The present invention describes a method for monitoring the health of physical assets using digital twins, which is carried out through the following stages: diagnosis of the assets, prioritizing those with the greatest relevance in the production process; gathering information about a specific asset through topographical tools and CAD models for the construction of a digital twin; installation of sensors in the asset in the optimal locations determined by the digital twin, in order to monitor critical variables; processing and analysis of the data obtained by the sensors, comparing them with the expected behavior of the asset, using a local or cloud server; issuing automatic alerts when the data exceeds predefined thresholds; and finally, the display of the information on a real-time dashboard, allowing users to monitor the status of the asset and the alerts issued.

[0018] The invention also contemplates a system for monitoring the health of physical assets using digital twins, which includes: configurable sensors designed to measure critical variables of the asset, which are installed in said asset; a data acquisition system that collects the information in real time and transmits it to a server, either local or in the cloud, for processing; a digital twin generated from CAD or 3D models of the asset, which reproduces the expected behavior of the monitored asset; processing software that compares the collected data with the modeled behaviors and generates alerts if the measured values ​​​​are outside the predefined thresholds; and finally, a control system that displays the information in real time on a control panel, showing the current status of the asset and the alerts generated.

[0019] According to one embodiment, the method comprises creating a digital twin of the asset using an internal API that transforms CAD models into a dynamic system. This digital twin is used to simulate the asset's behavior under operational conditions and serves as a reference for evaluating the information obtained by the sensors installed on the asset. Furthermore, the digital twin allows the parameters of the installed sensors to be dynamically adjusted based on the data received, thereby improving monitoring accuracy.

[0020] According to another embodiment, the system and method include the installation of various types of configurable sensors that measure critical variables such as acceleration, tilt, temperature, displacement, and deformation of the asset. These sensors are highly accurate, and their installation is optimized based on the location determined by the digital twin, maximizing data collection. The sensors are connected to a data acquisition system capable of simultaneously processing multiple variables in real time. The collected information is compared with models generated by the digital twin to identify potential deviations from the asset's expected behavior.

[0021] In another embodiment, both the method and the system include artificial intelligence software that analyzes historical and real-time data to identify anomalous patterns and potential operational failures. When sensor data exceeds predefined thresholds, the system generates automatic alerts that are transmitted to users through various channels, including email, SMS, WhatsApp, and Microsoft Teams. These alerts allow users to make quick decisions to mitigate risks and ensure the operational continuity of the asset.

[0022] In another embodiment, the system incorporates a redundancy system in data acquisition to ensure continuous monitoring. In the event of a failure of the primary server, the sensors continue collecting and transmitting information through a backup system, ensuring uninterrupted monitoring of the asset. This redundancy system is key to maintaining monitoring integrity, especially in critical assets where failures can have serious operational consequences. The redundancy system may include data storage in at least one memory.

[0023] In another embodiment, the system includes a dashboard that provides real-time visualization of the asset's condition, presenting variables such as vibration frequencies, inclinations, and displacements. Additionally, the dashboard provides access to a history of past events that have exceeded established thresholds, providing a comprehensive analysis of the asset's behavior over time. The system facilitates downloading reports and detailed analysis of events for long-term decision-making. BRIEF DESCRIPTION OF THE FIGURES

[0024] As part of the present invention, the following representative figures are presented, which teach preferred embodiments and, therefore, should not be considered as limiting the definition of the technical subject matter disclosed. Fig. 1: Diagram of essential components of the technology. Fig. 2: Representative diagram of the technology implementation process. Fig. 3: Shows a first example of the information display. Fig. 4: Shows a second example of information display. EXAMPLE No. 1 OF TECHNOLOGY

[0025] The technology includes, for example, the implementation of the following process: a) Asset inspection and diagnosis is performed in coordination with the end user, and their feedback is essential for prioritizing the assets to be deployed. This process includes, in addition to reviewing available information such as plans, calculation reports, and plant operating diagrams, a visit to the end user's facilities to observe how the asset has performed to date. b) Data collection is performed by operator users, seeking the information necessary to build the asset's digital twin.This search is not limited by the documents that the end user can provide. If necessary, a physical survey of the asset is performed using topographic tools, including levels, total stations, RTK GNLASS GPS equipment, and LIDAR laser scanners. This stage of the technology concludes with a 3D model in Computer-Aided Design (CAD) software that allows for the construction of the digital twin at a later stage. The digital twin that is built is for the specific purposes of monitoring and contains all the details necessary to assess its integrity. The digital twin also allows for identifying the locations where sensor installation will provide the most information possible, and it also allows for combining security criteria for sensor installation in this location assessment. Furthermore, the models help generate the thresholds that will generate asset integrity alerts, which is where the data that will allow information processing to determine the criticality of the event will be generated. The digital twin is built from the collected information (CAD file) where an Application Programming Interface (API) allows rapid transition from the three-dimensional model to the digital twin for monitoring purposes. This API builds the model from the geometry and generates a dynamic system that represents the asset's behavior for monitoring purposes. This dynamic system that is developed (digital twin) allows analyzing whether the information received from the sensors is compatible with the expected behavior of the asset and, if not, generating the corresponding alerts. In addition, it allows maintaining a complete history of the asset even though the instruments are installed at specific points. particular, because the modeling is performed on the entire asset and not just at the sensor installation locations. c) The technology uses commercial sensors, such as sensors that have been custom-developed through the construction of embedded systems. These latter provide extremely precise information for monitoring purposes and are also designed to facilitate and make their installation in the field safer. Sensors can be of any nature and any type of variable that allows for assessing the asset's integrity. This information is acquired regardless of the sensor type, whether electrical signals, images, videos, or information publicly available on the internet. Sensors can be wireless or directly connected to the data acquisition server, which is installed on the asset and whose purpose is to acquire data, search for relevant patterns, and send it for further processing in the cloud or on a local server. The data acquisition system is characterized by its ability to acquire any data, whether these are time-varying signals such as acceleration, displacement, deformation, temperature, wind speed, electric current, voltage, among others, and videos or images or any other information, linking all of these into a single time stamp, which allows for much more in-depth analysis and faster and more accurate diagnoses. The acquired information is stored in a non-sequential database. This database stores the data received online, as well as event data (where calibrated thresholds in the digital twin have been exceeded). This database includes information from all sensors, cameras, or any other data source, both before and after the event. Having pre-event information is essential for a good diagnosis. The data acquisition system's capacity is one of the technology's key benefits, allowing for monitoring any asset—both structural monitoring, which is preferred, and assets for monitoring gas abatement equipment, mining truck exhaust monitoring, remote testing, and more. The versatility of the technology even allows, in a local example, to monitor the flow of the river that passes over a bridge by means of queries to the DGA database (information available on the internet), and with that it allows to correlate, for example, the inclination of a bridge with the flow of the river in case of a major flood. d) The data is received in the cloud or on the local server and as a first step a verification is made that the received data do not constitute a false positive, if not the processing continues. As a first step, the requirements included in codes or design specifications are evaluated in relation to magnitude limitations in the sensors according to their location, subsequently an analysis is carried out to determine their criticality, this by behavioral analysis and comparison of expected behaviors with the digital twin. All of this processing is done autonomously in the cloud without human intervention. Finally, the procedure assesses the asset's criticality and generates a 4-level indicator to detail its status: green (OK); yellow (caution); orange (requires assistance); and red (requires field modifications to avoid a short-term interruption in asset operations). (The colors vary depending on the colors adopted by end users for different alert levels.) This can be implemented using software. a) The control panel or dashboard initially displays real-time information, which may be analog or digital electrical signals corresponding to physical, chemical, or environmental variables, images, videos, among others. It then graphically presents the results of the processing obtained in the previous stage and, of course, the alert level according to the colors indicated above.The dashboard can also display information from previous events, generating an event history that can be consulted in relation to the intensity of previous events compared to the one analyzed. The dashboard allows reports to be downloaded based on dates entered by the user and raw data collected by the sensors.

[0026] Through modeling and the information received, it is possible to monitor structural, thermal, and other production-related behavior. Exceeding these thresholds triggers advanced processing of the information received, and in this context, the technology implements different levels of information distribution. Table 1 shows an example of the information associated with the technology in question:

[0027] Table 1

[0028] In general, an example of the technology implementation process corresponds to what is shown in Figure 2, which shows a representative diagram of said process. EXAMPLE No. 2 OF TECHNOLOGY

[0029] In an analyzed mining operation, 7 assets were identified for monitoring, which correspond to: - Stockpile and Tripper Car: asset whose instrumentation is oriented to monitoring the stresses in the structural elements, considering the location of the tripper car. - Mill: This belongs to the line of essential assets and therefore, instrumentation is proposed that is capable of monitoring the vibrations generated by the mill as well as the stresses transferred to the structure. - East Belt: The belt presents complications due to friction with metallic elements. The objective of monitoring in this case is to monitor how the stresses are redistributed among the elements due to the impact of the reduction in section produced by this friction, and in general, to monitor the stresses of the structure. - Correa Poniente: It is the continuation of the previous strap that presents the same complications. - Dome: The instrumentation for this Dome is related to monitoring the redistribution of stresses resulting from the collapse due to undermining of one of the foundations. - Crusher: Monitoring in crushers involves tracking operational variables such as vibration frequencies and their impact on the structure and monitoring the stresses produced in structural elements. - Feeder: The feeder is an essential asset that presents a lot of vibrations, and the purpose of the instrumentation is to observe how these impact the stresses of the elements and how this condition could generate problems in the elements in the future due to the occurrence of fatigue.

[0030] The technology, in this example, is implemented with a physical asset monitoring system that contains the following elements: - Power Supply System: Based on a UPS, the system can operate for at least four hours after a power outage, ensuring the continuity of measurement services. The power supply has a minimal noise level so as not to disrupt measurements of structural variables. - Acquisition Controller: A server is considered to control the system's acquisition. This server will use at least an Intel® N5105 processor with 8 GB of memory. RAM. This server connects to cloud services via a dual-SIM modem (allowing it to work with two telephone companies). All servers have a secure access protocol for any necessary maintenance or modifications. - Data acquisition: The acquisition system is composed of analog-to-digital lobit converters, capable of reaching sampling rates of up to 2000 Hz per channel. - Sensors: High-sensitivity accelerometers are used, also used to measure inclinations. The accelerometer is housed in an aluminum housing coated with corrosion-resistant polyurethane paint. At least four accelerometers and three strain sensors are included in each asset. Exceptionally, and where necessary, displacement sensors will be considered. The characteristics of these sensors are presented in Table 2.

[0031] Table 2

[0032] In order to maintain the highest level of reliability in the sensors, the following technologies have been implemented: - Accelerometers: Considering the very low current through the conductors (450 pA), the cable length will not affect the measurement when using high-impedance acquisition equipment. To avoid ground loops, fully shielded paired cables are used. Cable lengths of no more than 40 m are expected in this project. - Strain Sensor: These sensors are installed using a signal conditioner located less than 3 meters from each sensor. This allows the signal magnitude to be amplified and transformed into a 4-20mA signal, enabling smooth conduction over long distances. - Displacement Sensor: It is considered a laser displacement sensor which allows measuring with an accuracy of ±1mm up to a distance of 500m.

[0033] These sensors will monitor the following parameters: - Statistics on Measured Variables: An online record will be kept of the peak, mean, standard deviation, and correlation of each of the received signals. This data can be used to track changes that can be used to detect a change in the asset's condition. - Turns: From the accelerometer readings, the angle of inclination of the 2 axes of interest is determined, both instantly and in moving average values. - Asset Dynamics: Accelerometer information will also be used to measure the asset's dynamic variables, specifically monitoring Reiher-Meister intensities and vibration frequencies.

[0034] Then, in this example the technology is implemented by the following: - Online Monitoring: In addition to sending data generated from specific conditions (through threshold exceedances), the system is capable of sending online data related to the variables being measured by the sensors, which is stored in a database. - Definition of Alert Thresholds: If necessary, thresholds can be defined to allow event acquisition. These will trigger a system response. For all events, information is organized according to the configuration required for subsequent processing. Thresholds are defined in the engineering phase based on statistical behaviors such as peaks, averages, and correlations. Once the event is declared, data is collected from a specific time before the event was generated and then processed in the cloud. The system can determine whether the event is a false positive or a real event. Threshold monitoring will also be based on their trend, that is, how close the threshold is to a given event if the asset's conditions persist for a period of time.In addition, test events are implemented at least twice a day, which allows for more continuous information on the different variables. - Display Interface (Dashboard): This is considered a display interface tailored to the end user. The following elements are considered part of this interface: 1. Log in with email and password. 2. Considering that the instrumentation is for a single asset, an information screen will immediately appear, as shown in Figures 3 and 4. The images above correspond to the information collected in the cloud. Regarding the system capacity, there is no limit to the number of assets to monitor; in this case, it will be developed for just one. 3. A set of polar graphs that will display KPIs from different sensors of an instrumentation node where they coexist with different colors. 4. A summary table with the most important data acquired. 5. A list of the latest alerts received, indicating the date, time, and reason that triggered the alert. 6. A list indicating the latest automatic system checks (autocheck) and the response obtained (OK, or requires attention). 7. Option to request a historical data report delivered by email. 8. An automatic alert system via email or SMS

[0035] Additionally, one modality of the technology implements a trained artificial intelligence that aims to achieve rapid event diagnosis, identification of anomalies in measurement parameters, and spatial coordination of RGB and thermal cameras.

[0036] Finally, below is an example implementation associated with a user interface deployed on the dashboard, which in this example is called the RhenÜX Visualization Interface. The following elements are considered part of this interface: - Access system: The main objective of the RhenÜX access system is to be able to manage access to the platform only to authorized users, safeguarding the information contained. - Overview: The main purpose of the RhenÜX home screen is to have an overview of the monitored assets as well as the latest occurrences on assets. - Asset View: The main objective of the RhenÜX asset screen is to have an overview of a monitored asset in addition to the latest occurrences, online data and other relevant key indicators. - Alert View: The main purpose of the RhenÜX alert screen is to be able to manage a particular alert and take into account everything that happened during the period that the alert was active. - Event View: The main objective of the RhenÜX event display is to have a more detailed view of a particular event, in order to make decisions based on the available information and the predefined criteria for each particular signal. - History View: The main objective of the RhenÜX History screen is to be able to manage documents and events in an intuitive, orderly manner following a temporal logic.

[0037] In addition to the core elements, the platform offers other external features such as automatic reporting based on information available online, and alerts via email, WhatsApp, or Microsoft Teams for continuous monitoring. EXAMPLE No. 3 OF TECHNOLOGY

[0038] In another implementation modality, the technology is applied to thermal and hydrological monitoring of physical assets through the integration of specialized sensors and the use of digital twins adapted to these variables.

[0039] In an industrial production center that involves heat transfer processes, a heat exchanger unit is instrumented using surface temperature sensors and non-contact infrared sensors. The exchanger's digital twin is built from its CAD model and calibrated using historical operating thermal conditions, allowing simulation of the asset's behavior under different thermal gradients. Thermal information is collected by sensors connected to the data acquisition system and processed to identify deviations from the simulated model. When a predefined thermal threshold is exceeded, an automatic alert is generated and the condition is displayed on the control panel.

[0040] Additionally, the technology is used to monitor structures located near waterways. In these cases, the system integrates level and flow data from external sources, such as the public database of the General Directorate of Water (DGA), and correlates them with structural variables measured at the facility, such as inclinations and displacements. This integration makes it possible to anticipate risks associated with flooding or undermining, and to dynamically adapt alert thresholds based on real-time hydrological conditions.

Claims

CLAIMS 1. A method for monitoring the health of physical assets using digital twins, comprising the following stages: - diagnose assets, prioritizing the most critical ones based on their value in a production process; - gather information about an asset using surveying tools and CAD models to build a digital twin; - install sensors on the asset according to locations determined by the digital twin, in order to monitor relevant variables; - process and analyze data received from sensors, comparing them with the expected behavior according to the digital twin, using a local or cloud server for data processing; - issue automatic alerts when data exceeds predefined thresholds, sending notifications to users; and - Display information on a real-time dashboard, allowing users to view asset status and alerts.

2. The method of claim 1, wherein the digital twin of the asset is built using an internal API that allows the CAD models to be transformed into a dynamic system to simulate the behavior of the asset and determine the optimal locations for the installation of the sensors.

3. The method of claim 1 or 2, wherein the sensors installed on the asset are of different types, including acceleration, deformation, inclination, temperature, displacement and other types of sensors specifically designed to capture relevant variables according to the criticality of the asset.

4. The method of claim 1 or 3, wherein the digital twin enables dynamic calibration of the installed sensors, adjusting their operating parameters based on the comparison between the measured data and the expected behavior of the asset.

5. The method of any one of claims 1 to 4, wherein the step of processing and analyzing the data includes using artificial intelligence software to identify patterns. anomalies or potential failures in the asset's behavior, comparing historical and real-time data.

6. The method of any one of claims 1 to 5, wherein the information displayed on the control panel includes processed data and real-time graphics presenting vibration frequencies, inclinations, correlations between variables and a history of events that have exceeded defined thresholds.

7. The method of any of claims 1 to 6, wherein the automatic alerts are generated when the data exceeds the defined thresholds, sending notifications through one or more channels such as email, SMS, WhatsApp and / or Microsoft Teams.

8. The method of claim 5 or 7, wherein a predictive behavioral analysis of the asset is performed using the digital twin, which allows anticipating critical events based on the operating conditions and history of the asset.

9. The method of any of claims 1 to 8, further comprising the step of acquiring data through a redundancy system, wherein the sensors continue to collect and transmit data through a backup system in the event of server failure, ensuring continuity of asset monitoring.

10. A system for monitoring the health of physical assets using digital twins, comprising: - configurable sensors to measure relevant variables of an asset, installed in said asset; - a data acquisition system, which collects information from sensors in real time and transmits it to a local or cloud server for data processing; - a digital twin, generated from 3D or CAD models of the asset, which reproduces the expected behavior of the monitored asset; - processing software, which compares sensor data with the behavior modeled by the digital twin and generates alerts when it detects values ​​outside predefined thresholds; - a control system, which displays information in real time through a control panel, showing the status of the asset and the alerts generated.

11. The system of claim 10, wherein the configurable sensors include acceleration, tilt, strain, temperature, and displacement sensors designed to capture critical variables of the asset.

12. The system of claim 10 or 11, wherein the data acquisition system is comprised of high-precision analog-to-digital converters capable of sampling multiple variables simultaneously at a rate of up to 2000 Hz per channel.

13. The system of any of claims 10 to 12, wherein the digital twin is generated by an internal API that converts the CAD models of the asset into a dynamic system that simulates the structural and operational behavior of the asset.

14. The system of any of claims 10 to 13, wherein the server is configured to analyze in real time the information obtained from the sensors and compare it with the behavior modeled by the digital twin.

15. The system of any of claims 10 to 14, wherein the control system displays the information through the control panel that includes real-time graphics, event histories and alerts generated according to the operating conditions of the asset.

16. The system of any of claims 10 to 15, comprising a redundancy system, wherein the sensors are configured to continue transmitting data through a backup system in the event of a server failure, ensuring continuity of asset monitoring.

17. The system of any of claims 10 to 16, wherein alerts generated by the control system are automatically transmitted to users through one or more channels, when predefined operating thresholds are exceeded.

18. The system of any of claims 10 to 17, wherein the control panel displays vibration frequencies, inclinations, displacements and other parameters of the asset, correlated with the alerts generated in real time.

Images