Method, apparatus, system, device, and medium for analyzing field element status
Patent Information
- Application Number
- PCT/CN2023/109665
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-07-27
- Publication Date
- 2025-01-30
AI Technical Summary
The prior art is difficult to realize real-time data acquisition and analysis of field elements in industrial scenarios without affecting the security and stability of existing systems, especially in systems that need to comply with high security standards.
By installing image capture devices in front of the Field Element controller, capture the LED indicator data on the display, and use the edge computing device for data filtering and analysis, generate time sequence data, transmit it to the storage or interface, and realize the state Element state state Forecast and analysis.
Realized real -time analysis and prediction of the Field Element status, avoiding the needs of the hardware and software upgrades of the Legacy system, reducing the risk of re -approval of security certificates, and reducing the needs of on -site visits and additional sensor installation, thereby reducing the reduction Cost and fault risk.
Smart Images

Figure CN2023109665_30012025_PF_FP_ABST
Abstract
Description
Method, apparatus, system, device, and medium for analyzing field element statusTECHNICAL FIELD
[0001] The present disclosure relates to a method, apparatus, device, and medium for analyzing field element’s behavior and status, in industrial field.BACKGROUND
[0002] Legacy computer systems, such as previous generation electronic interlockings are not prepared for remote access to capture real time data from connected field elements. The limited processing power and data transmission capacity of such systems does not allow for additional services without the risk of disrupting current operations. Field elements, such as point machines in mobility industry, and electric machines in factory automation fields and other on-site moving elements usually are connected with such legacy computer or controller or element controller especially with LEDs as indicators of working condition or alarm of the moving or said field elements etc.,
[0003] If such field elements and their legacy controllers form parts of critical infrastructure, for example infrastructure with safety integrity level SIL3 / 4, it may even not be allowed to change the hardware or software, which would render related safety certificates invalid. Hence, there is little to no chance to develop “Digital Twins” and offer predictive maintenance concepts for the installed base. However, replace a whole infrastructure’s legacy computer systems could be very time consuming and costly. This leads to that user is bound to preventive and corrective maintenance with limited system availability and penalties for sudden service disruptions.
[0004] The problem of not having remote access to real time field element data in legacy controller / computer systems has not been sufficiently solved yet.
[0005] In general, element control computers detect element-or internal failures and generate an alarm message to be observed in the control center. But such alarms happen after a problem occurs and give no indication to avoid the problem upfront and don’t give details of the root cause, which could be found in log files within the element controller only.
[0006] There are solutions to add remote access to log-files stored in the system, provided that a SW-upgrade and renewal of the safety assessment is possible. But the log-files only provide aggregated and rather static samples of the operational status and do not reflect the dynamic, real-time behavior of actuators, sensors, and other connected computers on low level.
[0007] Another option to solve the problem is to install additional sensors with data communication in the field to monitor field elements. However, this option needs access to protected customer sites, which is costly and introduces another mode of failure to the system if sensors break. In many cases, the customer sites’ system shall continuously run. If system operations get disrupted, it is often due to (sporadically) failing field elements. Without any remote access, experts from user or supplier side must rush to the physical location of the computer where the field elements are connected, which can be very remote and try to get the log-files manually out of the system to analyze them at user’s office or even worse later at the supplier’s lab. The turn-around time from failure to root-cause reporting is between days and weeks with additional measurements in the fields if the root-cause can be identified at all.
[0008] To avoid sudden failures, users usually are over-maintaining the system and keeping a large stock of spare parts for immediate repair and hence having massive asset-and operational costs.SUMMARY
[0009] The solve above listed problems, embodiments of the present disclosure propose a method, apparatus, system, electronic device and medium for analyzing and predicting field element status.
[0010] In a first aspect, a method for predicting field element behavior is provided. The method comprising:
[0011] receiving image data captured by at least one image-capturing device , wherein said image data include images and / or video data with at least one indicator, indicating behavior of at least one field element ,
[0012] generating filtered data based on said image data as time series data,
[0013] transmitting said filtered data in a memory storage for storing or,
[0014] transmitting said filtered data to one or more interfaces.
[0015] With being non-intrusive, the disclosure provides advantages over known solutions: the HW / SW of the legacy system doesn’t need to be upgraded and hence avoids the reapproval of the safety certificate for mission critical systems. The current running system doesn’ t risk interruptions from alteration works or CPU overload with additional analytics tasks in unexpected situations. And there is then with the disclosure no need to install any additional sensors and data transmissions in the field to get field element data and hence avoid side access, additional costs and problems from sensor or data transmission failures.
[0016] According to embodiments of the disclosure, before transmitting, detecting and / or predicting the status information of said at least one indicator according to the filtered data, and generating information of field element’s behavior. With such embodiments, edge computing device could on-site output analysis and prediction of the field elements. With help of LED or indicators configuration files, the function could be further enhanced. Without such configuration files, the indicators behavior and link with field controller and field element can also be observed and learnt.
[0017] According to embodiments of the disclosure, the method according further comprising:
[0018] detecting and / or predicting the status information of said at least one indicator according to the filtered data, generating information of field element’s behavior.
[0019] Such detecting and / or prediction can be implemented on IoT platform, with support of machine learning model or neural networks, especially when data is transmitted with interfaces to IoT platforms.
[0020] According to embodiments of the disclosure, wherein detecting and / or predicting the status information of said at least one indicator, comprises:
[0021] analyzing said filtered data by machine learning or neural network model, wherein said machine learning or neural network model is trained with filtered data, alarm information of said field element and environment data .
[0022] The method according to any one of embodiments, further comprises:
[0023] transmitting the filtered data in real time to cloud based IoT platforms.
[0024] The method according to any one of embodiments, further comprises:
[0025] output analyzed result to an external representation of said field element.
[0026] By giving access to low-level / HW related data, embodiments of disclosure provide the base for a Digital Twin concept: the preprocessed video data allow the graphical reconstruction of the entire dynamic behavior of the computer at any given time.
[0027] Field elements like electric motors (e.g. in point machines or barrier drives) and sensors (e.g. axle counters and track circuits) as well as cable based data transmission systems degrade slowly over time and their dynamic behavior and degrading is well reflected in the gradually changing of on / off times (digital information) of the LED’s on the element controller boards.
[0028] The disclosure provides not only field element data but also all other indications from CPU-and communication boards in real time which allows for a complete process view of system behavior on HW-level. The disclosure allows for advanced maintenance concepts and offers related cost advantages and KPI benefits:
[0029] The real-time analysis of degrading field elements allows for predictive maintenance concepts. A reduction of maintenance related costs by stretching the regular maintenance cycles is saved, which for example in mobility industry for a typical metro line over thousand hours per year can be saved. This can be scaled to city / country or even global level with thousands of existing lines in operation and it is not limited to Mobility but can be used for any older field element controllers.
[0030] In a second aspect, an apparatus for analyzing element’s status is provided, comprising:
[0031] a receiving module, configured to receive image data captured by at least one image-capturing device,
[0032] wherein said image data include images and / or video data with at least one indicator of at least one field element,
[0033] a generating module, configured to filter data based on said image data with time series data,
[0034] a transmitting module, configured to transmit said filtered data to a memory storage OR to transmit said filtered data to one or more interfaces.
[0035] In a third aspect, a system for analyzing status of field elements is provided, comprising:
[0036] an image capturing device, configured to capture image data of at least one indicator for indicating at least one field element’s working status,
[0037] an edge computing device, configured to generate filtered data based on image data received from said image capturing device, with time series data,
[0038] one or more interfaces communicated with said edge computing device.
[0039] In a fourth aspect, a method of training machine learning model or neural network model, comprise:
[0040] labelling the filtered data using the alarm information of the said element and
[0041] generating alarm related events,
[0042] correlating, the alarm related events with environmental data,
[0043] classifying alarm related events and output anomaly detection and / or
[0044] performing regressing analysis for alarm related events, which then predicting anomaly of field element.
[0045] In a fifth aspect, an electronic device is provided. The electronic device comprising a processor and a memory, wherein an application program executable by the processor is stored in the memory for causing the processor to execute a method for predicting carbon emissions from an industrial process as described in any of the above.
[0046] In a sixth aspect, a computer-readable medium comprising computer-readable instructions stored thereon is provided, wherein the computer-readable instructions, when executed by a processor, implement a method for building knowledge graph or a method for predicting carbon emissions from an industrial process as described in any of the above.
[0047] In a seventh aspect, a computer program product comprising a computer program, when the computer program is executed by a processor for executing a method for predicting carbon emissions from an industrial process as described in any of the above.BRIEF DESCRIPTION OF THE DRAWINGS
[0048] In order to make technical solutions of examples of the present disclosure clearer, accompanying drawings to be used in description of the examples will be simply introduced hereinafter. Obviously, the accompanying drawings to be described hereinafter are only some examples of the present disclosure. Those skilled in the art may obtain other drawings according to these accompanying drawings without creative labor.
[0049] FIG. 1 is a schematic diagram of the system for analyzing and predicting elements’s tatus according to embodiments of the present disclosure.
[0050] FIG. 2 is a flowchart of a method for analyzing elements’ behavior status according to embodiments of the present disclosure.
[0051] FIG. 3 is a schematic diagram of apparatus for analyzing elements’ behavior status according to an embodiment of the disclosure.
[0052] FIG. 4 is a schematic diagram of process for determining a prediction model of a type of field element according to an embodiment of the present disclosure.
[0053] FIG. 5 is a schematic diagram of electronic device according to embodiments of the present disclosure.
[0054] List of reference numbers: DETAILED DESCRIPTION
[0055] In order to make the purpose, technical scheme and advantages of the disclosure clearer, the following examples are given to further explain the disclosure in detail.
[0056] In order to be concise and intuitive in description, the scheme of the disclosure is described below by describing several representative embodiments. Many details in the embodiments are only used to help understand the scheme of the disclosure. However, it is obvious that the technical scheme of the disclosure can be realized without being limited to these details. In order to avoid unnecessarily blurring the scheme of the disclosure, some embodiments are not described in detail, but only the framework is given. Hereinafter, "including" refers to "including but not limited to" , "according to... " refers to "at least according to..., but not limited to... " . Due to the language habits, when the number of an element is not specifically indicated below, it means that the element can be one or more, or can be understood as at least one.
[0057] Independent of the grammatical term usage, individuals with male, female or other gender identities are included within the term.
[0058] In any industry, most older systems implicitly provide field element data as many field element control computer or controller systems have LEDs on their PCB front panels to indicate the current status and level of data processing from connected field elements.
[0059] Although such LEDs or other measure of indicators were not intended for predictive maintenance, the pattern and timing of flickering of LEDs for example may represent the actual behavior of connected field elements and adjacent connected computers. Especially field elements like electric motors (e.g., in point machines (转辙机) or barrier drives and other motion or moving elements) and sensors (e.g. axle counters (计轴器) and track circuits (轨道电路) as well as cable based data transmission systems degrade slowly over time before they break.
[0060] The dynamic behavior and degrading of such elements are well reflected in the gradually changing of on / off times (digital information) of the LED’s on the element controller boards. Of course, if the indicators are in other format, such as bars, flickering points shown in the screen, the changing colors and on / off times could also reflect the dynamic behavior and degrading of those elements.
[0061] Taking the mobility industry for example, the toggle time of a point machine represents the time to switch the point from one position to the other and depends on the mechanical resistance of the switching blades in the track which is an indicator for necessary maintenance or failure in the future. The activation of the motor of the point machine might be indicated by an LED on the motor controller board for the time of activity. But this low-level processing of field element data information is hidden on the controller board. The flickering pattern of LED’s have not been used for advanced analytics. According to the embodiments of the disclosure, such flickering pattern of LED is used for analyzing the condition of the point machine, and similarly for other control elements in other industry.
[0062] Fig. 1 show embodiments of the disclosure with a non-intrusive detect and predict system for analyzing and predicting field element behavior, especially anomaly. The system includes an edge computing system 100, and a neural network model 400. The embodiments of the edge computing system 100 includes an image capturing device 10, for example a wide-lens video high speed camera or camera array. The image capturing device, or for example high speed camera 10 can be set in front of a legacy system 200 such as an element controller 200. Although it is not shown in the figure, the camera 10 can be set e.g., inside of the door of a computer cabinet to capture the display 210 with LEDs 230 and / or other visual indicators (if there is any) at the front panel of the element controller 200. In embodiments of the disclosures, there can be a group of element controllers 200, which are connected to different field elements 300. Those element controllers 200 each have their panel with LEDs for example and are communicated with the control center 500. It can be understood that LEDs is just one example of indicators on panel of the controller. Indicators 230 can also be bars and other visualized graphic user interface indicator on screens or panels. Below embodiments will take LEDs as example.
[0063] Indicators, or such as LEDs in the embodiments flicker according to the status of the connected executor 300. For example, one LED turns from green to red indicating failure of the field element 300, or several LEDs start to flash quickly indicating moving part’s speed increasing of the field element 300. In embodiments of mobility industry, the LEDs flickering of the controller for point machine could indicate that the movement of the blades become slower. Such LEDs’ flash, flicker and color changing pattern can be predefined with LED configuration files 40 in an edge computing device 30 which is communicated with the camera 10.
[0064] In embodiments of this disclosure, element controllers 200 are communicated with several different field elements 300. The element controllers 200 can be arranged one by one or stack together. Different indicators 230 on the panel of corresponding controllers 200 reflect the status or warn a failure or anomaly status of corresponding field elements 300. The video camera or camera array 10 record or capture the images / videos of the LEDs 230 on the panel 210 of the element controller 200. The image data or video data is transmitted to an edge computing device 30. The edge computing device 30 can be an edge computing unit or a micro controller. The edge computing device 30 analyze the raw video data transmitted from the camera 10 and generates a filtered data stream with time series data, which means the edge computing device 30 adds time stamp to the video data. In the edge computing device 30, LED configuration file 40 can be installed. With the LED configuration file 40, the LEDs do not need to be artificially detected in the video stream as there are system configuration files 40 with the static position of the LEDs and their corresponding field element available. With the configuration file 40, static positions of LEDs and link of those LEDs with field element is available for the edge computing device 30 to process with video data to analyze and provide relations of LEDs, field element and status (such as warnings) of field elements as well as stamped timing of the status happening time. It is also understandable that in case there is no LED configuration file 40, the edge controller could find the LED positions themselves by video analytics. The filtered data represents each status change of LEDs as well as alarm-LED filtering. For example, alarm LED of one element controller 200 flickered in red color. This alarm event is captured by the camera 10 and generate video stream transmitted to the edge computing device 30. The edge computing device 30 with configuration file 40 link the position of the alarm LED to a specific field element and add time stamp to the video. With such processing, the raw process video is able to tell at which timing, which field element has failure or gets alarm information.
[0065] Depending on the computing ability of the controller 30 a machine learning model can be applied to detect abnormal or slowly changing pattern in LED behavior and generate warnings for maintainers directly on the system 100, for example using a separate screen or laptop connected to the controller 30 or directly on the edge device 30 to show such warning information (not shown in the FIG. 1) . The warning information can be transmitted to the on-site maintainers via the interfaces 63, 67 for example directly on the edge device.
[0066] By processing the captured image of the LEDs with the configuration files 40 and added time stamp, the controller 30 generates filtered information, which is transmitted to one or more interfaces 61and 67 or be stored in the memory storage 50 connected to the controller 30 and been transmitted to one or more interfaces 63 and 65. It can be understood that, although there are several different interfaces shown for example in FIG. 1, those interfaces can be integrated into one or more interfaces and connected to for example storage 50, IoT platform, on-site computer via different kinds of networks, according to the application scenarios.
[0067] With the memory storage 50, all processed data stream of LEDs behaviors can be stored in the storage 50, for transmission on demand, for example via cRSP (secure file access via remote service platform) . This makes limitations of data volume for the timeframe of interest possible. This access could be triggered on dispatching level once an alarm was registered. In this way, retrospective remote analytics is realized. The memory storage 50 can be a T-byte Flash drive with IP interface 63.
[0068] In embodiments, high speed camera is used as the image capturing device. For example, for filtered data volume: 400 LEDs (20*20 Array) per system in a defined (not random) configuration can be distinguished and captured with a low resolution (10k Pixel) but high speed (e.g. 16000 fps=frame per second) video camera (a) . Raw video stream (160Mbyte / s) processing on an edge device 30 is generating a filtered data stream with time series data representing each status change of a single LED with <100 microsecond time stamp accuracy. If each LED changes two times per second, 4 *400 data points per sec with accurate time stamps (e.g. 20 byte per change) will be generated. This results in manageable 32k Byte / sdata transmission and 2~3G Byte data per day or 1-2T Byte per year. With the edge computing system 100 as shown in FIG 1, the video stream is processed into manageable volume to store in the memory storage 50 for future analysis in the edge device 100 or transmitted to cloud-based IoT platform for predicting.
[0069] The filtered data stream can be transmitted in real time via secure data interface e.g., 61 in FIG. 1 to IoT platforms for further data analytics. Of course, those interfaces could also give access to connect laptops for on-site access to processed data in real-time or to the storage 50. As shown in FIG. 1 the edge computing system 100 could provide several interfaces for connecting with IoT platforms, for connecting a laptop for on-site access to processed data, and for connecting with storage 50 with both IoT platforms, on site computers and touch displays via Wi-Fi, 5G, Ethernet, and other Industrial networks.
[0070] A predict system, according to embodiments of the disclosure is enabled to provide predicting ability based on the filtered data stream. A trained neural network model implemented in the IoT platform could analyze the filtered data stream and predict the status of the field element 300.
[0071] As shown in FIG. 1, the filtered data with time stamp information is transmitted to the IoT platform, where the predicting machine learning model 400 is applied. The predicting machine learning model 400 is also communicating with the control center 500 for the field elements 300, wherein the element controller’s alarm is sent to the control center 500, and control center 500 is able to communicate such alarms to the classifier module 410 of the machine learning 400. The classifier module 410 , which includes a labeling module 412 and a correlation module 414, receives alarm messages from the control center 500 as well as alarm messages extracted from the filtered data with time stamp from interfaces 61, 63. With the labeling module 412, the alarm information is configured to trigger the labeling of “alarm event” autonomously, then for example the event filter data such as 10 second prior to the timing of the alarm is classified as a “alarm event” . In meanwhile, environmental data 401 is also fed to the classifier module 410. The preprocessed filtered data stream from edge computing system 100 is correlated with the high context environmental data 401 from non-safety critical operation control or / and maintenance environmental systems connected or communicated with the field element 300. Environment data for example includes such as temperature, humidity, maintenance logfiles. Those environment data can be transmitted from control center 400 and / or from internet through APIs for example. The labeled “alarm event” is then correlated with the environment data in the correlating module 414, so that the correlated data can be further input into the learning model 420 for training. The learning module 420 could comprise a scenario classification module 422 and a scenario regression module 424. The labeled correlated data allows for supervised machine learning or deep learning, either with association rule learning or sequential pattern mining to do scenario classification for anomaly detection and making scenario regression for predictions about future mis-behavior of the field elements 300 in real time. By inputting the correlating data into the machine learning scenario classification module 422, the machine learning model could realize anomaly detection for field element 300. By inputting the correlating data into the machine learning scenario regression module 424, the machine learning model could realize anomaly prediction for field element 300.
[0072] The trigger to analyze data of particular field elements for a given time frame and the return of the results for past, present or future can be further realized via API (Application Programming Interface) so that an external user front end application 810 can poll the data for visualization in a dashboard or to animate a static model (e.g., BIM) and form a “Digital Twin” 800 of the field element 300 in the Industrial Metaverse. As also shown in FIG 1, the machine learning model 400 processed output data is input through API to digital twin 800 of field elements 300. The output data provide status data with time stamp and machine learning detection and prediction, which enables the digital twin with status features and allows users to check the status of the field element 300 for past, present, and future.
[0073] It can be understood that the machine learning model 400 can also be a neural network model 400. With machine learning and neural networks, it is able to filter the massive amount of field data and correlate them with real time data from surrounding systems to make useful predictions and detect anomalies.
[0074] FIG. 2 is a flowchart of a method for analyzing and predicting the behavior status of field element 300 according to embodiments of the present disclosure. As shown in Figure 1, the method comprises:
[0075] Step 110: receiving image data captured by at least one image-capturing device 10. The edge computing device 30 receives said image data, which include images and / or video data with at least one indicator 230 of at least one field controller connected with at least one field element 300. As shown in FIG 1, for example, the edge computing device 30 receives the video stream from the high-speed camera 10, wherein the LEDs and other GUI indicators’ behavior is recorded in the video or images.
[0076] Step 120: generating filtered data based on said image data with time series data. The edge computing device 30 analyze the raw video data (image stream data) transmitted from the camera 10 and generates a filtered data stream with time series data by adding time stamp for example. The filtered data represents each status change of LEDs, attached with time series.
[0077] Step 131: transmit and stored in the memory storage 50 OR
[0078] Step 132: to one or more interfaces 61~67. As explained in the embodiments, those interfaces 61~67 can be connected with the storage 50 for storing the filtered data for later use and analysis. The interfaces can also be used to connect with on-site computer for on-site analysis and / or transmitted to IoT platform 400.
[0079] In embodiments of the disclosure, before transmitting the filtered data stream in the memory storage 50 or before transmitting to the interfaces, the edge computing device 30 is configured to:
[0080] Step 171: detecting the status information of said at least one indicator 230 according to the filtered data, and
[0081] Step 173: generating information of element’s behavior.
[0082] Then the elements’ behavior information is further as in Step 131 and 132 transmitted to the memory storage 50 or the interfaces 61~67. As explained for embodiments as shown in Figure 1, such step can be realized by applying machine learning model or neural network to the edge computing device 30 for analysis and predicting of the field elements’ behavior, especially for anomaly detection.
[0083] The information of element’s behavior as processed by edge computing device 30 or alternatively the filtered information directly transmitted trough the interfaces to IoT platforms can be further processes as in below steps: Step 140: detecting the status information of said at least one indicator 230 according to the filtered data, and in Step 150: generating information of element’s behavior.
[0084] As explained for the embodiments of the disclosure, in step 140, the status information of said at least one indicator, comprises the alarm of the element, warnings and working conditions of the element. In such embodiments, the edge computing device 30 can be provided by indicator configuration file 40, for such comparison.
[0085] In embodiments of the disclosure, detecting the status information of said at least one indicator, comprises analyzing said filtered data by neural network model and predicting the behavior of said elements 300, wherein said neural network model is trained with indicators’ pattern and behavior data. When the edge computing device 30 has sufficient computing capacity, the neural network or machine learning model can be applied to the edge computing device 30, and / or it can be in the IoT platform 400 for cloud computing.
[0086] FIG. 3 is a schematic diagram of apparatus for analyzing elements’ behavior status according to an embodiment of the disclosure. The apparatus comprising:
[0087] a receiving module 110, configured to receive image data captured by at least one image-capturing device 10,
[0088] wherein said image data include images and / or video data with at least one indicator 230 of at least one field element 200,
[0089] a generating module 120, configured to filter data based on said image data with time series data, a transmitting module 130, configured to transmit said filtered data to a memory storage 50 OR transmit said filtered data to one or more interfaces 61~67.
[0090] FIG. 4 is a schematic diagram of process for determining a machine learning model or neural network model according to an embodiment of the present disclosure. The neural network model can be trained in following methods, which comprises:
[0091] labelling (S201) the filtered data using the alarm information of the said element 30 and
[0092] generating (S202) labeled alarm events,
[0093] correlating (S203) , the labeled alarm events with environmental data 401,
[0094] classifying (S204) alarm events and output anomaly detection and / or
[0095] performing regressing analysis (S205) for alarm events, which then predicting (S206) anomaly of field element (300) .
[0096] Such neural network model or machine learning model can be implemented on the IoT platform for analysis and predicting of the field element’s behavior. It can also by applied in the edge computing device 30.
[0097] Embodiments of the present disclosure also provide an electronic device with processor-memory architecture. FIG. 5 is a structural diagram of an electronic device according to an embodiment of the present disclosure.
[0098] As shown in FIG. 5, the electronic device 600 comprises a processor 601, a memory 602, and a computer program stored in the memory 602 and executable on the processor 601. When the computer program is executed by the processor 601, any one of the above-mentioned methods for analyzing and field element’s behavior is implemented. The memory 602 may specifically be implemented as various storage medium such as Electrically Erasable Programmable Read-Only Memory (EEPROM) , Flash memory (Flash memory) , Programmable Program Read-Only Memory (PROM) . The processor 601 may be implemented to include one or more central processing units or one or more field programmable gate arrays, wherein the field programmable gate arrays integrate one or more central processing unit cores. Specifically, the central processing unit or central processing unit core may be implemented as a CPU or MCU or DSP, and so on.
[0099] It should be noted that not all steps and modules in the above-mentioned processes and structural diagrams are necessary, and some steps or modules may be omitted according to actual needs. The execution logic of each step is not fixed and can be adjusted as needed. The division of each module is only to facilitate the description of the functional division used. In actual implementation, a module can be implemented by multiple modules, and the functions of multiple modules can also be implemented by the same module. These modules can be in the same device or in a different device.
[0100] The hardware modules in various embodiments may be implemented mechanically or electronically. For example, a hardware module may include specially designed permanent circuits or logic devices (e.g., special-purpose processors, such as FPGAs or ASICs) to perform specific operations. Hardware modules may also include programmable logic devices or circuits temporarily configured by software (e.g., including general-purpose processors or other programmable processors) for performing operations. As for the specific mechanical method, or a dedicated permanent circuit, or a temporarily configured circuit (e.g., configured by software) to realize the hardware module, it can be decided according to cost and time considerations.
[0101] The above descriptions are merely preferred embodiments of the present disclosure and are not intended to limit the protection scope of the present disclosure. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present disclosure shall be included within the protection scope of the present disclosure.
Claims
1.A method for analyzing field element (300) ’s status, comprising:receiving (S110) image data captured by at least one image-capturing device (10) , wherein said image data include images and / or video data with at least one indicator (230) , indicating behavior of at least one field element (300) ,generating (S120) filtered data based on said image data as time series data,transmitting (S131) said filtered data in a memory storage (50) for storing or,transmitting (S132) said filtered data to one or more interfaces (61~67) .2.The method according to claim 1, characterized in thatbefore transmitting (S131, S132) , detecting and / or predicting (S171) the status information of said at least one indicator (230) according to the filtered data, andgenerating (S173) information of field element (300) ’s behavior.3.The method according to claim 1, comprising:detecting and / or predicting (S140) the status information of said at least one indicator (230) according to the filtered data,generating (S150) information of field element (300) ’s behavior.4.The method according to claims 2 or 3, wherein detecting and / or predicting (S140, S171) the status information of said at least one indicator, comprises:analyzing said filtered data by machine learning or neural network model, wherein said machine learning or neural network model is trained with filtered data, alarm information of said field element (300) and environment data (401) .5.The method according to any one of claims 1 to 4, further comprises:transmitting (S160) the filtered data in real time to cloud based IoT platforms (400) .6.The method according to claim 4, further comprises:output analyzed result to an external representation of said field element (300) .7.An apparatus for analyzing element’s status, comprising:a receiving module (110) , configured to receive image data captured by at least one image-capturing device (10) , wherein said image data include images and / or video data with at least one indicator (230) of at least one field element (200) ,a generating module (120) , configured to filter data based on said image data with time series data,a transmitting module (130) , configured to transmit said filtered data to a memory storage (50) OR to transmit said filtered data to one or more interfaces (61~67) .8.A system (100) for analyzing status of field elements (300) , comprising:an image capturing device (10) , configured to capture image data of at least one indicator (230) for indicating at least one field element (200) ’s working status,an edge computing device (30) , configured to generate filtered data based on image data received from said image capturing device, with time series data,one or more interfaces (61~67) communicated with said edge computing device (30) .9.The system (100) according to claim 8, wherein:said edge computing device (30) is configured to detect and / or predict the changing pattern of said at least one indicator (230) according to the filtered data, andgenerate information of element (300) ’s status.10.The system (100) according to claim 8 or 9, wherein:said edge computing device (30) is configured to communicate with one or more interface (61~67) , wherein,said filtered data is transmitted to a IoT platform and analyzed by a machine learning model or neural network model (400) to detect and predict the status of said elements (300) .11.The system (100) according to anyone of claims 8-10, wherein: said machine learning model or neural network model (400) is configured to comprise:a labelling module (412) , which receives alarm information of the said element (300) and label the filtered data using the alarm information and generate labeled alarm events,a correlating module (414) , which correlates the labeled alarm events with environmental data (401) ,a scenario classification module (422) , which classifies alarm events for anomaly detection,a scenario regression module (424) , which predict anomaly of field element (300) based on regression analysis for alarm events.12.The system (100) according to claim 11, wherein said system (100) is configured with API (600) for receiving the request from and transmitting output of said machine learning model or neural network model (400) to an external representation (800) of said field elements.13.Method of training machine learning model or neural network model (400) , comprise:labelling (S201) the filtered data using the alarm information of the said element (30) andgenerating (S202) alarm related events,correlating (S203) , the alarm related events with environmental data (401) ,classifying (S204) alarm related events and output anomaly detection and / orperforming regressing analysis (S205) for alarm related events, which then predicting (S206) anomaly of field element (300) .14.An electronic device (600) , comprising a processor (601) and a memory (602) , wherein an application program executable by the processor (601) is stored in the memory (602) for causing the processor (601) to execute a method for analyzing status of field element according to any one of claims 1-6.15.A computer-readable medium comprising computer-readable instructions stored thereon, wherein the computer-readable instructions for executing a method for analyzing status of field element according to any one of claims 1-6.16.A computer program product comprising a computer program, upon the computer program is executed by a processor for executing a method for analyzing status of field element according to any one of claims 1-6.
Citation Information
Patent Citations
Monitoring activities in industrial sites
US10896331B2
Video streaming user interface with data from multiple sources
US11158177B1
Remote surveillance
US20060179463A1
Virtual reality and augmented reality for industrial automation
US20180131907A1
Augmented reality safety automation zone system and method
US20190147655A1