Search

EP-3973364-B1 - SYSTEM AND METHOD FOR DETERMINING A HEALTH CONDITION AND AN ANOMALY OF AN EQUIPMENT

EP3973364B1EP 3973364 B1EP3973364 B1EP 3973364B1EP-3973364-B1

Inventors

  • RAO, Prashanth Belur Gururaja

Dates

Publication Date
20260506
Application Date
20191025

Claims (9)

  1. A system for determining a health condition and an anomaly of an equipment (104), wherein the system comprises: a plurality of sensors (106A-N) that is attached to the equipment (104), wherein the plurality of sensors (106A-N) is configured to sense information associated with a plurality of parameters of the equipment (104); a field device (110) that is communicatively connected to the plurality of sensors (106A-N), wherein the plurality of sensors (106A-N) is configured to wirelessly communicate the sensor data to the field device (110); a camera unit (108) that is configured to capture visual data associated with the equipment being analyzed, wherein the camera unit (108) is configured to wirelessly communicate the captured images to the field device (110); and a server (112) that is communicatively connected to the field device (110) for receiving the sensor data and the visual data associated with the equipment (104), wherein the server employs a machine learning model (410) for determining a health condition or an anomaly of the equipment (104), and wherein said server (112) comprises: a database (114) configured to store (i) the sensor data and (ii) the visual data, wherein the sensor data comprises values of the plurality of parameters associated with the equipment (104), and wherein the visual data comprises at least one of (i) a plurality of images of the equipment or (ii) a plurality of videos of the equipment (104); and a processor configured to execute a set of modules, wherein the set of modules comprises (i) a fault detection module (202) configured to process the sensor data to determine a fault or the health condition of the equipment (104), (ii) an image processing module (204) configured to be trained to detect an anomaly in the equipment (104) by processing the visual data associated with the equipment (104), and (iii) a report generation module (206) configured to automatically generate a health report (212) based on the health condition of the equipment (104) and the anomaly detected in the equipment (104), wherein the fault detection module (202) is configured to communicate with the sensor data and the image data stored in the database (114) and to make use of the machine learning model (410) to determine an anomaly if there is a fault, wherein the machine learning model (410) is configured to get trained periodically by image processing module (204) that is configured to analyse the visual data and determines the type of fault accordingly based on historical data available, wherein the visual data comprises photographs of the equipment (104) captured at different angle, wherein the machine learning model (410) is configured to combine the photographs of the equipment (104) to get a more complete idea of a a position of the equipment (104) in a physical world, wherein the machine learning model (410) is configured to analyse the visual data and to determine fault based on the historical data, wherein the health report (212) comprises at least one of (i) a location of the equipment (104), (ii) an observation based on the sensor data and the visual data, (iii) a recommendation to rectify a fault or an anomaly detected in the equipment (104), (iv) a criticality level of the fault or the anomaly detected in the equipment (104) or (v) a reference to industry standards, and wherein the health report (212) is communicated to client device.
  2. The system as claimed in claim 1, wherein the processor comprises a work order module (208) configured to generate a work order (210) when initiated by the client device, wherein the work order (210) comprises a request to determine a health condition of the equipment (104) and an anomaly in the equipment (104).
  3. The system as claimed in claim 1, wherein the sensor data and the visual data of the equipment (104) are tagged with at least one of (i) a customer identifier, (ii) a work order (210) identifier, (iii) a location identifier, (iv) a facility identifier, (v) a floor identifier, (vi) an equipment (104) type, (vii) an anomaly type or (vii) an image identifier or a video identifier.
  4. The system as claimed in claim 1, wherein the machine learning model (410) is configured to identify a cause that is attributed to the anomaly, wherein the cause of the anomaly is attributed to a design of the equipment (104), installation of the equipment (104), maintenance of the equipment (104) and operating conditions of the equipment (104).
  5. The system as claimed in claim 1, wherein the field device (110) is configured to communicate (i) the sensor data and (ii) the visual data to the server (112) in a plurality of stages, wherein after each of the plurality of stages, the server (112) is configured to generate an alert to the field device, wherein the alert comprises data regarding consistency and relevance of (i) the sensor data and (ii) the visual data.
  6. A method for determining the health condition and an anomaly of equipment (104), wherein the method comprises the steps of sensing, using a plurality of sensors (106A-N), information associated with a plurality of parameters of the equipment (104), wherein the plurality of sensors (106AN) is attached to the equipment (104); communicating, using the plurality of sensors (106A-N), the sensor data to a field device (110), wherein the field device (110) is communicatively connected to the plurality of sensors (106A-N); capturing, using a camera unit (108), visual data associated with the equipment being analysed, wherein the camera unit (108) is configured to wirelessly communicate the captured images to the field device (110); receiving, using a server (112) that is communicatively connected to the field device (110), the sensor data and the visual data associated with the equipment (104), wherein the server employs a machine learning model (410) for determining a health condition or an anomaly of the equipment (104); generating a database (114), in a server (112), with (i) the sensor data and (ii) the visual data, wherein the sensor data comprises values of the plurality of parameters associated with the equipment (104), and wherein the visual data comprises at least one of (i) a plurality of images of the equipment (104) or (ii) a plurality of videos of the equipment (104); processing, using a fault detection module (202) of the server (112), the sensor data to determine a fault or the health condition of the equipment (104); communicating, using the fault detection module (202), with the sensor data and the image data stored in the database (114) and making use of the machine learning model (410) to determine an anomaly if there is a fault, wherein the machine learning model (410) gets trained periodically by image processing module (204) that analyses the visual data and determines the type of fault accordingly based on historical data available; training an image processing module (204) of the server (112) to detect an anomaly in the equipment (104) by processing the visual data, wherein the visual data comprises photographs of the equipment (104) captured at different angle, wherein the machine learning model (410) combines the photograph of the equipment (104) to get a more complete idea of a position of the equipment (104) in a physical world, wherein the machine learning model (410) analyses the visual data and determines a fault based on the historical data; and generating, using a report generating module (206) of the server (112), a report based on the health condition of the equipment (104) and the anomaly detected in the equipment (104), wherein the health report (212) comprises at least one of (i) a location of the equipment (104), (ii) an observation based on the sensor data and the visual data, (iii) a recommendation to rectify a fault or an anomaly detected in the equipment (104), (iv) a criticality level of the fault or the anomaly detected in the equipment (104) or (v) a reference to industry standards, and wherein the health report (212) is communicated to a client device.
  7. The method as claimed in claim 6, wherein the method comprises generating, using a work order module (208) of the server (112), a work order (210) when initiated by the client device, wherein the work order (210) comprises a request to determine a health condition of the equipment (104) and an anomaly in the equipment (104).
  8. The method as claimed in claim 6, wherein the method comprises tagging (i) the sensor data and (ii) the visual data with at least one of (i) a customer identifier, (ii) a work order (210) identifier, (iii) a location identifier, (iv) a facility identifier, (v) a floor identifier, (vi) an equipment (104) type, (vii) an anomaly type or (vii) an image identifier or a video identifier.
  9. The method as claimed in claim 6, wherein the method comprises communicating (i) the sensor data and (ii) the visual data to the server (112) in a plurality of stages from the field device (110), wherein after each of the plurality of stages, the server (112) generates an alert to the field device (110), wherein the alert comprises data regarding consistency and relevance of (i) the sensor data and (ii) the visual data.

Description

BACKGROUND Technical Field The embodiments herein generally relate to automation of analytical diagnostic tool, and more particularly relate to a system and method for determining the health condition and anomalies of an equipment for carrying out electrical and fire safety audits for detecting failures and also to predict possible corrosion in equipment. Description of the Related Art In industrial setups, there are very significant electrical equipments that play a major role in the work-flow of the industry. These equipments are manufactured while keeping in mind their reliability, integrity, durability, etc. However, these equipments are just man-made machines which are bound to fail at some point. The reason for failure is a wide spectrum of possibilities that cannot be determined with accuracy and efficiency by human effort. The reason for failure may include manufacturing defects, anomalies, non-adherence to the industry standards, non-ideal conditions of operation, etc. Traditionally, human intervention is required to monitor these equipments for anomalies and proper functioning. It remains impossible to efficiently monitor the equipments at all times and to predict the next point of failure. Typically, health reports are generated by human intervention where a lot of effort is required to articulate all the relevant information in a single document. This process is repetitive and traditionally requires a lot of human effort. Existing approaches have been developed to address the above problem and which involves the use of a field engineer or an expert who can interpret the data from the equipment directly with his expertise. This way of data collection is very time consuming and the accuracy of the data is dependent on the skill of the field engineer or the expert who is working to get the required data from the machine or the equipment that has been tested. The tendency of the human to evaluate or process the gathered information and generate a health report is based on various factors that tend to change as per time. US 2017/032281 A1 relates to a system that encompasses web portals for data capture, maintenance, quality assurance, and weld engineering at fabricator locations. The system facilitates real-time display of weld quality and equipment service predictions on various user interfaces and utilizes cloud computing, employing technologies such as REST API, Node.js, J2EE, and HTML5 and enables customization for fabricators while minimizing the need for labor-intensive algorithm customization and incorporates disparate data for machine learning model training, employing cloud services for efficient pre-processing of raw data. Further, US 2017/032281 A1 discloses specific system architectures for integration, assembly of data, and condition-based maintenance for enhancing predictive capabilities and optimizing maintenance procedure. US 2016/292846 A1 relates to a system that involves determining machine states using video and/ or audio analytics for optimizing worksite operations. Performance and diagnostic data are collected from various machines at the worksite, including payload, efficiency, productivity, fuel economy, and maintenance-related information. Each hauling machine is equipped with an onboard control module, operator interface module, and communication module to facilitate data transmission to a central station. The control module includes sensors distributed throughout the hauling machine, capturing data related to machine states such as load conditions, payload placement, machine position, and more. Data received by the control module is communicated to the central station through the communication module. Accordingly, there remains a need for a system and method which can monitor the equipment in real-time with improved accuracy without false observations. SUMMARY In view of the foregoing, the present invention provides a system for determining the health condition and an anomaly of an equipment according to claim 1. In some embodiments, the processor comprises a work order module that generates a work order when initiated by a client device, wherein the work order comprises a request to determine a health condition of the equipment and an anomaly in the equipment. In some embodiments, the sensor data and the visual data of the equipment are tagged with at least one of (i) a customer identifier, (ii) a work order identifier, (iii) a location identifier, (iv) a facility identifier, (v) a floor identifier, (vi) an equipment type, (vii) an anomaly type or (vii) an image identifier or a video identifier. In some embodiments, (i) the sensor data and (ii) the visual data are communicated from the field device to the server in a plurality of stages, wherein after each of the plurality of stages, the server generates an alert to the field device, wherein the alert comprises data regarding consistency and relevance of (i) the sensor data and (ii) the visual data. According to the inven