Periodic inspection of electrical installations

Abstract

The invention relates to a computer-implemented method for periodic inspection of an electrical device, in particular a field device (1), comprising receiving image data related to the electrical device, evaluating the image data to determine whether the electrical device is in a normal state, wherein the evaluation of the image data is based on machine learning and / or artificial intelligence, and generating a report indicating whether the electrical device is in the normal state. According to another aspect of the invention, a system for periodic inspection of an electrical device is presented. Furthermore, a computer program product for periodic inspection of an electrical device is presented.

Description

Technical Field

[0001] This invention relates to a computer-implemented method for periodically inspecting electrical installations (particularly field equipment). Furthermore, a system for periodically inspecting electrical installations is also proposed. According to another aspect of the invention, a computer program product for periodic inspection is provided. Background Technology

[0002] Electrical installations must be inspected regularly to ensure their safe operation. For example, visual inspection can be performed to identify certain defects (such as contamination or corrosion). Periodic inspection is a repetitive or periodically performed inspection. For example, DIN VDE 0105-100 specifies the procedures for conducting periodic inspections. Errors can occur during manual inspection. Some defects may be easily missed by the human eye. Furthermore, the judgment of human inspectors is subjective. For example, misjudgments may occur due to inattention or fatigue. Summary of the Invention

[0003] Therefore, an object of the present invention is to provide a method for technically inspecting electrical installations, wherein the method is objective and particularly reliable to ensure that the installations achieve a higher level of safety. The method should also be quick and easy for the user to implement. Another object of the present invention is to provide a system for inspecting electrical installations. Furthermore, a computer program product for inspecting electrical installations will be provided. This object is achieved by the method according to claim 1, the system according to claim 14, and the computer program product according to claim 15. The dependent claims relate to alternative embodiments of the invention.

[0004] It should be noted that the features listed in the independent and dependent claims can be combined in any way that is technically reasonable. This also applies to situations crossing claim class boundaries, even if one claim does not depend on another. The specification also features and specifically describes the invention in conjunction with the accompanying drawings. If numerical terms such as "first" or "second" are used in the claims or specification to describe elements, these numerical terms are used only to distinguish elements and do not indicate the order, total number, or other characteristics of the elements, unless otherwise expressly stated.

[0005] According to a first aspect of the invention, a computer-implemented method for periodically inspecting electrical installations (particularly field equipment) is proposed. The method includes the steps of: receiving image data associated with the electrical installation; evaluating the image data to determine whether the electrical installation is in a normal condition, wherein the evaluation of the image data is based on machine learning and / or artificial intelligence; and generating a report indicating whether the electrical installation is in a normal condition. This method automatically determines whether the electrical installation still meets technical requirements. With the aid of computer-aided evaluation, situations where electrical installations continue to operate, for example, due to subjective misjudgment, can be avoided, thereby achieving a higher safety standard.

[0006] This method can be performed specifically via a mobile terminal (e.g., a smartphone or tablet). However, it can also be performed via other data processing devices. For periodic inspection of electrical installations, a user can record images or videos of the installations, preferably using the camera of a mobile terminal. This generates image data, which is then evaluated according to the invention. Preferably, the evaluation is performed using methods from the fields of machine learning and / or artificial intelligence. For example, neural network-based methods can be used in the evaluation. However, other methods from the fields of machine learning and / or artificial intelligence, such as statistical methods, can also be used. According to the invention, a hybrid approach can also be used, meaning that different methods from the field of machine learning can be combined.

[0007] For the purposes of this invention, "electrical device" should be understood to mean any electrical equipment, and in particular, also includes electronic equipment. Electrical devices are particularly advantageously field devices. In process automation technology, field devices are frequently used to detect and / or influence process variables. Examples of such field devices include fill level measuring devices, limit level measuring instruments, and pressure measuring instruments with sensors that detect relevant process variables (i.e., fill level, limit level, or pressure). However, field devices may also have actuators, such as control valves. Field devices known in the prior art typically have a housing, sensors, and a control unit arranged within the housing.

[0008] Preferably, an assessment is performed to check whether the electrical device is in a normal working condition. According to the invention, a report is output stating whether the electrical device is in a normal working condition. In principle, the report can concisely state whether the electrical device is in a normal working condition. However, according to the invention, the report can also include additional information. According to the invention, the report can be a digital message (e.g., a notification in a program, email, or SMS). However, the report can also be output in document form (e.g., as a PDF file), or written to or output to a database. The report is preferably output via an application, particularly advantageously via an application executed on a mobile terminal using the method according to the invention.

[0009] The evaluation method according to the invention preferably uses a pre-trained model or an AI model. For example, the model can be in the form of a dataset containing weights from a neural network. The model is preferably trained to identify whether an electrical device is in a normal state based on image data. Preferably, the model is trained to work with different electrical devices. Furthermore, it is more advantageous if the model is trained to detect multiple faults in the electrical device. Preferably, the model is trained to classify the severity of one or more defects in the electrical device and preferably to infer whether the electrical device is in a normal state. The model can be trained using supervised learning in particular. For training, image data of the electrical devices is processed as input data, wherein each electrical device is assigned a label indicating whether it is in a normal state and / or whether a defect exists. This process can use image data associated with the electrical device or different types of electrical devices. According to the invention, the input data may also include type information of the electrical device. This enables the model to learn how to accurately evaluate different types of electrical devices.

[0010] When evaluating image data to determine whether an electrical installation is in proper condition, it is preferable to check for at least one safety-related defect and / or whether the installation conforms to current construction standards. In particular, it can be checked whether the installation is in proper condition conforming to DIN VDE 0105-100. Preferably, only visual inspection of the installation is performed, meaning only visual data is processed. According to the invention, characteristics of the installation that are not based on DIN VDE 0105-100 can also be checked. In particular, visual inspection can also be performed according to other standards. However, periodic inspections do not necessarily have to be performed according to standards.

[0011] According to an advantageous embodiment of the invention, the at least one safety-related defect is a defect caused by aging, wear, operation, and / or environmental conditions. For example, over time, electrical devices may develop defects due to the aging of materials or components. This can lead to malfunctions. Wear and tear may also occur, for example, due to the wear of moving parts. Furthermore, contamination may also occur, particularly under extreme environmental conditions, such as underwater or inside machinery. Preferably, the model on which the evaluation is based is trained to visually detect such defects in the electrical device.

[0012] According to an advantageous variation of the invention, the report identifies at least one defect in the electrical installation. The report may optionally include: information about the type of the at least one defect; information about the severity of the at least one defect; information about the location of the at least one defect in the electrical installation; recommendations for repairing the at least one defect; and / or recommendations for inspection of the electrical installation by a technician. According to the invention, the report may also contain additional information, tips, and / or instructions. According to an advantageous embodiment of the invention, the report is archived, and / or the data contained in the report is archived, for example, in a database. This allows for subsequent access to the results of periodic inspections.

[0013] The assessment preferably includes: an assessment of the readability of the nameplate of the electrical device; an assessment of the degree of corrosion of the electrical device's housing; an assessment of the degree of contamination at the process connection points and / or housing of the electrical device; and / or an assessment of the degree of physical damage to the electrical device. According to the invention, only one of these defects may be inspected, or multiple defects may be inspected. According to the invention, the presence of other defects may also be checked.

[0014] To assess whether an electrical device is in a normal state based on identified defects, these defects can be evaluated individually. For example, the severity can be determined for each defect. According to an advantageous embodiment, if at least one defect exceeds a predefined severity level, a message indicating that the electrical device is not in a normal state is displayed. According to an alternative embodiment of the invention, different types of defects are comprehensively evaluated to determine whether the electrical device is in a normal state. For example, an arithmetic mean of the severity levels assigned to different types of defects can be determined. Furthermore, different weights can be applied to different types of defects during the evaluation process.

[0015] Advantageously, the image data includes at least one digital photograph of the electrical device and / or at least one video recording of the electrical device. Video recording in the sense of this invention includes all types of moving image recordings, particularly digital moving image recordings. Depending on the application scenario, it may be necessary to generate either a digital photograph or a video recording. For example, a digital photograph may be characterized by a higher level of detail compared to a video recording, but this typically depends on the terminal used to capture the image data. Furthermore, certain defects may only be identifiable in video recordings, such as if the moving parts of the electrical device no longer move or only move within a limited range.

[0016] Advantageously, the method also includes receiving type information indicating the type of electrical device, wherein the type information is used as the basis for evaluating the image data. Therefore, type information is preferably used during the evaluation process as well. Type information can be, for example, the model number or serial number of the electrical device. In principle, this type information is information capable of identifying the type of electrical device. Preferably, inspection history is retrieved based on the type information. For example, this inspection history can be stored in a database. It can indicate which inspections were performed on a particular electrical device in the past and include information about any defects found. According to an optional embodiment of the invention, features of the electrical devices inspected during the evaluation of the image data are selected based on the inspection history.

[0017] Preferably, the type information is based on user-input data. For example, a user can input the model number of the electrical appliance into the terminal, or select the type of electrical appliance to be inspected on the terminal's display. Alternatively or supplementarily, type information can be extracted from image data, wherein the type information is determined based on the appearance of the electrical appliance, the content of the electrical appliance's nameplate, and / or a QR code. The QR code may in particular be a QR code affixed to the electrical appliance. The image data is then visually evaluated based on these characteristics. This evaluation may, in particular, be based on artificial intelligence and / or machine learning. For example, a nameplate can be identified and its content read. However, type information can also be determined using other methods, such as by evaluating a QR code. According to an advantageous embodiment of the invention, type information is determined before the evaluation of the image data begins.

[0018] The method preferably includes issuing a first instruction to record a digital photograph or video of the electrical device's nameplate. This preferably prompts a user to record image data including the electrical device's nameplate. Based on the nameplate, the type of electrical device can be determined. For example, the first instruction can be issued to a user of a mobile terminal. Advantageously, the user of the mobile terminal uses an application that executes the method according to the invention. The first instruction can be contained in, for example, a text message, an audio message, or a similar medium.

[0019] According to the invention, the method may further include issuing a second instruction for recording image data, wherein the second instruction contains information such as: whether a digital photograph or video should be recorded; which details of the electrical device should be recorded; at what distance from the electrical device should recording be performed; at what optical magnification should recording be performed; from what angle should recording be performed; and / or whether a flash or additional lighting is used for recording. Therefore, according to an advantageous embodiment of the invention, instructions can be given regarding the conditions and / or the manner in which image data is generated. This ensures that relevant details for assessing whether the electrical device is in a normal state are included in the image data. This ensures complete and high-quality data collection. For example, the second instruction can be issued to a user of a mobile terminal performing the method according to the invention. The second instruction can be included, for example, in a text message, audio message, or similar carrier.

[0020] According to the present invention, multiple second instructions can be issued sequentially. For example, the user can first be instructed to photograph specific details of an electrical device. Then, the user can be instructed to set a specific optical magnification. This prevents the user from receiving too many instructions at once, allowing the user to gradually adjust the recording process as necessary. According to the present invention, the method can be designed such that the next second instruction is issued only when the requirements of the previous second instruction are met. This is preferably achieved through continuous or real-time evaluation of image data, for example, recorded by a mobile terminal. This can also be achieved using methods from the fields of artificial intelligence or machine learning.

[0021] The method preferably further includes receiving preliminary image data related to the electrical device, wherein an examination is performed to determine whether the preliminary image data contains at least one deficiency that would hinder the assessment of whether the electrical device is in a normal working state, and wherein the examination is preferably based on machine learning and / or artificial intelligence. If the preliminary image data contains at least one deficiency, a message indicating the at least one deficiency is preferably issued. This allows it to be determined whether the electrical device is in a normal working state based on the image data. If it cannot be assessed, the message can be used, for example, to notify a user of a terminal performing the method according to the invention. The message can be, for example, a text message, an audio message, or a similar form.

[0022] The evaluation methods based on machine learning and / or artificial intelligence are preferably trained to perform the aforementioned checks. For example, supervised learning can be used to train a model to identify deficiencies in image data. Examples can be used to learn in which cases image data can be correctly evaluated and in which cases it cannot be correctly evaluated.

[0023] The inspection of deficiencies is preferably performed in real time. This means that the preliminary image data will be evaluated immediately, or at least as soon as possible. In this sense, the inspection does not begin after the preliminary image data has been fully recorded, but rather during the recording of digital photographs or videos. Preferably, feedback on detected deficiencies is sent directly to the user, for example, directly to the user of the terminal performing the method according to the invention. However, according to other embodiments of the invention, the user may specify when the preliminary image data is completed. Only then will the inspection of deficiencies be performed.

[0024] According to an advantageous embodiment of the invention, at least one deficiency includes: lack of view of the electrical device; recording the electrical device from an inappropriate distance; recording the electrical device at an inappropriate angle; recording the electrical device with an inappropriate flash or inappropriate lighting; and / or insufficient image sharpness in the preliminary image data. However, it should be understood that, according to the invention, other deficiencies in the image data can also be identified. This requires appropriate training of the artificial intelligence or machine learning method used.

[0025] According to another aspect of the invention, a system for periodically inspecting electrical installations (particularly field equipment) is provided. This system includes a data processing device configured to perform the methods described above. The system for periodic inspections advantageously comprises at least one computer. The system for periodic inspections can also be comprised of a mobile terminal (e.g., a smartphone or tablet). The system preferably includes a unit for capturing image data, particularly a camera.

[0026] According to the invention, the system for periodic inspections can also be comprised of other computer systems (e.g., servers). According to the invention, the server can be configured to communicate with a mobile terminal, for example, via a network, and receive image data captured by the mobile terminal, as well as optional other data, from the terminal. This may include, for example, user input. The computer system is advantageously configured to also send output data to the mobile terminal, such as reports indicating whether electrical equipment is in a normal state. The computer system is preferably configured to perform the evaluation of the image data, either entirely or partially. According to possible variations of the invention, partial evaluation can also be performed by other computing units. This also applies to other partial steps or aspects of the method. In particular, partial aspects of the method can be performed by the mobile terminal. According to an advantageous variation of the invention, the system for periodic inspections includes a computer system and a mobile terminal. The computer system is preferably linked to the mobile terminal via a network connection.

[0027] According to another aspect of the invention, a computer program product is provided, comprising instructions that, when executed by a computer, cause the computer to perform the methods described above. The computer program product is preferably executed by, or used for execution on, the aforementioned system. The computer program product is preferably an application program for a mobile terminal. However, it can also be software running on different computer systems. According to the invention, the computer program product may also have multiple parts, thereby enabling it to be executed in a distributed manner, for example, on the aforementioned computer system and the mobile terminal. Attached Figure Description

[0028] Other features and advantages of the invention will become apparent from the accompanying drawings, which are intended to illustrate the invention to those skilled in the art so that they can practice it. However, the drawings do not limit the invention. Nevertheless, the drawings illustrate advantageous aspects of the invention, and their features can (even alone) be used to specifically illustrate the invention to be protected. Figure 1 A schematic diagram illustrating a scenario in which field equipment is periodically inspected according to the method of the present invention is shown. Figure 2 A flowchart of the method according to the present invention is shown. Detailed Implementation

[0029] Figure 1 A schematic diagram illustrating a scenario of periodic inspection of field device 1 according to the method of the present invention is shown. Field device 1 is a fill level measuring device installed in container 2. Container 2 is filled with liquid 3, and its fill level is determined by field device 1 using a radar sensor. Periodic inspections must be performed at regular time intervals. These periodic inspections are used to identify any defects in field device 1.

[0030] For periodic inspections, a mobile terminal 4 is used; in this example, the mobile terminal is a smartphone. The smartphone is equipped with a camera 5, which can be used to record photos and videos. An application for performing periodic inspections according to the invention is installed on the smartphone. When the user launches the application, the application instructs the user to use the camera 5 to take photos and record videos of the field device 1. The application provides the user with various instructions on how to generate the recordings. These instructions inform the user whether to take photos or record videos. Furthermore, it informs the user which parts of the field device 1 should be recorded. If the relevant parts of the field device 1 are not visible in the installed state, the application can also prompt the user to remove the field device 1 from its container. The application also provides the user with information on the optical magnification to be used when photographing the field device 1. To capture even the finest details, the application also provides guidance on the angle from which to photograph the details of the field device 1. Finally, the application can also activate the smartphone's flash to enable photography in sufficient brightness.

[0031] As the user generates a recording, the application continues to provide feedback on the captured footage. For example, if the image is blurry or lacks close-up views, the application will notify the user. Recordings unsuitable for evaluation will be deleted by the application. Once the user has completed all the necessary recordings, the application will inform the user that no further recording is required.

[0032] The application now evaluates the image data generated during recording to determine whether field device 1 is in normal working order. Artificial intelligence is used to evaluate the image data, specifically a neural network. This neural network is specially trained to perform evaluations. During the evaluation of the image data, the neural network identifies the nameplate 6 of field device 1. The nameplate contains the model number of field device 1. Therefore, the exact type of field device 1 can be determined, and an evaluation can be performed based on this. As part of the evaluation, the application according to the invention checks whether field device 1 is contaminated or damaged. If it is contaminated or damaged, the degree of contamination or damage will also be determined. Based on this, the application evaluates whether field device 1 is in normal working order.

[0033] The application then generates a report, which the user can view within the application. This report indicates whether field device 1 is in normal working order. If any defects are found in field device 1, these defects are also recorded in the report. The report indicates the severity of each defect. Furthermore, the report indicates the location of the defect within field device 1. Additionally, the report includes suggestions on how the user can repair the defects (if possible). If the user is unable to repair the defects, the report recommends that field device 1 be inspected by a technician.

[0034] Figure 2A flowchart of the method according to the present invention is shown. First, in receiving step 7, the application receives image data related to the electrical device. Subsequently, in evaluation step 8, it is determined whether the electrical device is in a normal state, wherein the evaluation of the image data is based on machine learning and / or artificial intelligence. Finally, in generating step 9, a report indicating whether the electrical device is in a normal state is generated.

[0035] According to the present invention, the above steps can be used in different application scenarios. Three exemplary application scenarios of the method are described below.

[0036] According to the first application scenario, pressure sensors are periodically inspected in a chemical plant. The inspection includes checking for corrosion on the pressure sensor housing, contamination at the process connection points of the pressure sensor, and the legibility of the nameplate. All of the above steps can be performed as part of a visual inspection, thus allowing the use of the method according to the invention. Users open the application according to the invention on their smartphones, scan the barcode of the pressure sensor, and are then guided through the inspection process by the application. The application prompts the user to take photos from specific angles (e.g., front view, side view, close-up of the connection points). The application analyzes the image data obtained in this way and identifies corrosion spots on the housing and difficult-to-read model markings. The application then generates a report, records these anomalies, and provides remedial recommendations, such as corrosion protection or replacement of the pressure sensor.

[0037] According to the second application scenario, sensors are regularly inspected in food processing plants. Hygiene integrity checks are performed on temperature sensors in dairy plants. In this case, it is particularly important that the temperature sensor housing is free of deposits or residues. The application according to the invention assists the user in capturing high-resolution images of the process connections and housing of the temperature sensor. During image evaluation, the application detects residues and identifies them as potential sources of contamination. The report generated by the application recommends cleaning and re-inspection before the temperature sensor is put back into use.

[0038] According to the third application scenario, visual inspection of level sensors is performed in petrochemical plants. Since harsh environments are expected to accelerate wear, physical damage inspection of the level sensors is conducted as part of routine checks. The user uses the application according to the invention and captures a video showing the entire sensor and its connections. The application analyzes the video and marks deformed process connection points. The application generates a report recording the exact location of the deformation and recommends a detailed inspection by a technician.

[0039] The method described herein is characterized by its ability to objectively assess electrical installations. This improves the safety of electrical installations. The reports generated according to this invention are advantageous because they enable complete recording and traceability of periodic inspections. Furthermore, this method saves time compared to manually generating inspection reports. For example, electronically generated inspection results can be integrated into existing maintenance systems or equipment databases for long-term condition monitoring. List of reference numerals 1. On-site equipment 2 containers 3. Liquid 4. Mobile terminals 5 cameras 6. Nameplate 7 Receiving Steps 8. Evaluation Steps 9. Generation Steps

Claims

1. A computer-implemented method for periodically inspecting electrical installations, particularly field devices (1), the method comprising: - Receive image data related to the electrical device; - Evaluate the image data to determine whether the electrical device is in a normal state, wherein the evaluation of the image data is based on machine learning and / or artificial intelligence; and - Generate a report indicating whether the electrical device is in a normal condition.

2. The computer-implemented method according to claim 1, Its features are, When evaluating the image data to determine whether the electrical installation is in a normal condition, check whether the electrical installation has at least one safety-related defect, and / or whether the electrical installation meets the current construction standards for the electrical installation.

3. The computer-implemented method according to claim 2, Its features are, The at least one safety-related defect is caused by aging, wear, operation, and / or environmental conditions.

4. The computer-implemented method according to any one of claims 2 or 3, Its features are, The report identifies at least one defect in the electrical installation, wherein the report may optionally include: - Information regarding the type of the at least one defect; - Information regarding the severity of the at least one defect; - Information regarding the location of the at least one defect in the electrical installation; - Recommendations for fixing the aforementioned defects; and / or - Recommendation to have the electrical equipment inspected by a technician.

5. The computer-implemented method according to any one of the preceding claims, Its features are, The evaluation of the image data includes: - An assessment of the readability of the nameplate (6) of the electrical device; - Assessment of the degree of corrosion of the housing of the electrical device; - An assessment of the degree of contamination at the process connection points and / or the housing of the electrical device; and / or - Assessment of the extent of physical damage to the electrical device.

6. The computer-implemented method according to any one of the preceding claims, Its features are, The image data includes at least one digital photograph of the electrical device and / or at least one video recording of the electrical device.

7. The computer-implemented method according to any one of the preceding claims, Its features are, The method further includes: Receive type information indicating the type of the electrical device, wherein the type information is used as the basis for evaluating the image data.

8. The computer-implemented method according to claim 7, Its features are, The type information is based on user-input data.

9. The computer-implemented method according to claim 7, Its features are, The type information is extracted from the image data, wherein the type information is determined based on the appearance of the electrical device, the contents of the nameplate (6) of the electrical device, and / or the QR code.

10. The computer-implemented method according to claim 9, Its features are, The method further includes: - Issue a first instruction to record a digital photograph or video of the nameplate (6) of the electrical device.

11. The computer-implemented method according to any one of the preceding claims, Its features are, The method further includes issuing a second instruction for recording the image data, wherein the second instruction contains the following information: - Should we record digital photos or videos? - What details of the electrical device should be recorded by photographing? - The distance from the electrical device should be recorded; - What optical magnification should be used for recording? - From what perspective should the recording be conducted; and / or - Whether to use a flash or additional lighting for recording.

12. The computer-implemented method according to any one of the preceding claims, Its features are, The method further includes: - Receive preliminary image data related to the electrical device; - Check the preliminary image data for at least one deficiency that would hinder the assessment of whether the electrical device is in a normal working condition, wherein the check is preferably performed based on machine learning and / or artificial intelligence; and - If the preliminary image data contains the at least one deficiency, then a message indicating the at least one deficiency is issued.

13. The computer-implemented method according to claim 12, Its features are, The at least one deficiency includes: - A view of the electrical device is missing; - Record the electrical device from an inappropriate distance; - The electrical device was recorded at an inappropriate angle; - Record the electrical device using an inappropriate flash or inappropriate lighting; and / or - The image clarity of the preliminary image data is insufficient.

14. A system for periodically inspecting electrical installations, particularly field devices (1), the system including a data processing device configured to perform the method according to any one of claims 1 to 13.

15. A computer program product comprising instructions that, when executed by a computer, cause the computer to perform the method according to any one of claims 1 to 13.

Images