Method, apparatus, device, and medium for diagnosing machining tool operations
An LLM-based diagnostic model for machining tools automates log interpretation, addressing the challenges of complex machining tool logs with efficient and accurate diagnostics.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- SIEMENS AG
- Filing Date
- 2024-12-27
- Publication Date
- 2026-07-02
AI Technical Summary
The complexity and volume of machining tool operation logs make it difficult for experts to quickly and accurately identify issues, leading to time-consuming and error-prone manual diagnostics.
A method utilizing an LLM-based diagnostic model trained on historical and standard operation logs to automatically interpret and generate log entries, providing real-time diagnostic feedback and suggesting corrective actions.
Significantly reduces the need for specialized human expertise and speeds up the diagnostic process by automatically identifying discrepancies and offering immediate corrective actions.
Smart Images

Figure CN2024143448_02072026_PF_FP_ABST
Abstract
Description
Method, apparatus, device, and medium for diagnosing machining tool operationsFIELD
[0001] The present disclosure relates to the technical field of industrial digitalization, in particular to a method, apparatus, device, and medium for diagnosing machining tool operations.BACKGROUND
[0002] In industrial settings, operation log files are crucial for diagnosing issues, optimizing operations, and ensuring smooth machine performance. Operation log files on machining tools are extracted and interpreted. This procedure requires significant expertise to decode and understand the data. However, the sheer volume and complexity of these logs make it difficult for even experienced expert to quickly and accurately identify problems or deviations from expected operating procedures. It’s not only time-consuming but also prone to human error, which can lead to incomplete or inaccurate interpretations.SUMMARY
[0003] Embodiments of the present disclosure propose a method, apparatus, device, and medium for diagnosing machining tool operations.
[0004] In a first aspect, there is provided a method for diagnosing machining tool operations, comprising:
[0005] obtaining historical operation logs and standard operation procedure logs, the standard operation procedure logs is generated from standard operations based on manuals;
[0006] training an LLM-based operation diagnostic model with the historical operation logs and standard operation procedure logs;
[0007] obtaining current operation logs;
[0008] diagnosing current operation corresponding to the current operation logs by the operation diagnostic model using the current operation logs.
[0009] Therefore, the present disclosure utilizes an LLM-based diagnostic model to automatically interpret and generate log entries, significantly reducing the need for specialized human expertise and speeding up the diagnostic process. Besides, the present disclosure offers real-time diagnostic feedback by comparing user-generated logs with standard operation procedures, immediately identifying discrepancies and suggesting corrective actions.
[0010] In an example, training an LLM-based operation diagnostic model with the historical operation logs and standard operation procedure logs comprising:
[0011] preprocessing the historical operation logs;
[0012] utilizing the preprocessed historical operation logs to learn and generate sequence of logs by unsupervised learning;
[0013] receiving labels from user for the sequence of logs and training the operation diagnostic model with the labels and the sequence of logs by supervised learning.
[0014] In an example, utilizing the preprocessed historical operation logs to learn and generate sequence of logs by unsupervised learning comprising:
[0015] utilizing the preprocessed historical operation logs, by a sequence-to-sequence model, to learn and generate sequence of logs by unsupervised learning.
[0016] In an example, the method further comprising:
[0017] setting up a test set;
[0018] generating an operation script, by the operation diagnostic model, with the test set;
[0019] optimizing the diagnostic model with the test set and the operation script.
[0020] In an example, the method further comprising:
[0021] generating current operation corresponding to the current operation logs;
[0022] calling an HMI playback interface to complete playback of the current operation;
[0023] detecting operation error in the current operation;
[0024] marking location and cause of the error operation based on standard operation.
[0025] In an example, the method further comprising:
[0026] detecting operation error in the current operation;
[0027] providing a step-by-step guidance for user to correct the error operation.
[0028] In a second aspect, there is provided an apparatus for diagnosing machining tool operations, comprising:
[0029] a first obtaining module, configured to obtain historical operation logs and standard operation procedure logs, the standard operation procedure logs is generated from standard operations based on manuals;
[0030] a training module, configured to train an LLM-based operation diagnostic model with the historical operation logs and standard operation procedure logs;
[0031] a second obtaining module, configured to obtain current operation logs;
[0032] a diagnosing module, configured to diagnose current operation corresponding to the current operation logs by the operation diagnostic model using the current operation logs.
[0033] In a third aspect, there is provided an 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 identifying building block as described in any of the above.
[0034] In a fourth aspect, there is provided a computer-readable medium comprising computer-readable instructions stored thereon is provided, wherein the computer-readable instructions for executing a method for identifying building block as described in any of the above.
[0035] In a fifth aspect, there is provided a computer program product comprising a computer program, when the computer program is executed by a processor for executing a method for identifying building block as described in any of the above.BRIEF DESCRIPTION OF THE DRAWINGS
[0036] 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.
[0037] Fig. 1 is an exemplary flowchart of a method for diagnosing machining tool operations according to an embodiment of the present disclosure.
[0038] Fig. 2 is an exemplary schematic diagram of a system for diagnosing machining tool operations according to an embodiment of the present disclosure.
[0039] Fig. 3 is an exemplary schematic diagram of a process for diagnosing machining tool operations according to an embodiment of the present disclosure.
[0040] Fig. 4 is an exemplary schematic diagram of an apparatus for diagnosing machining tool operations according to an embodiment of the present disclosure.
[0041] Fig. 5 is an exemplary structural diagram of an electronic device according to an embodiment of the present disclosure.
[0042] List of reference numbers:
[0043] 100 method
[0044] 110-140 step
[0045] 21 user
[0046] 22 machine tool
[0047] 23 LLM-based operation diagnostic model
[0048] 24 scenario
[0049] 25 scenario
[0050] 26 scenario
[0051] 201 standard operations
[0052] 202 standard operation procedure logs
[0053] 203 historical operation logs
[0054] 204 current operation logs
[0055] 301-308 steps
[0056] 401 first obtaining module
[0057] 402 training module
[0058] 403 second obtaining module
[0059] 404 diagnosing module
[0060] 500 electronic device
[0061] 510 processor
[0062] 520 memoryDETAILED DESCRIPTION
[0063] To make the purpose, technical scheme, and advantages of the disclosure clearer, the following examples are given to further explain the disclosure in detail. Nouns and pronouns related to people in this patent application are not limited to specific gender.
[0064] 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. 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... " . 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.
[0065] The present disclosure provides a method for diagnosing machining tool operations. Machining tool can be CNC machine tool. Fig. 1 is an exemplary flowchart of a method 100 for diagnosing machining tool operations according to an embodiment of the present disclosure. As shown in Fig. 1, method 100 comprises:
[0066] step 110, obtaining historical operation logs and standard operation procedure logs, the standard operation procedure logs is generated from standard operations based on manuals;
[0067] Historical operation refers to machining tool usage in the past, including both successful operations and those that resulted in errors or suboptimal outcomes. Historical operation logs are records of historical operation including actions, events, and parameters. These logs typically include timestamps, user inputs, machine states, and error messages. These logs can cover various operational scenarios to ensure diversity and comprehensiveness. Historical operation logs can be stored locally on the machine tool or centrally on the cloud for analysis.
[0068] Standard operations can be obtained from operation manual. Standard operations in the operation manual can serve as a reference point for identifying deviations in actual operations. Performing standard operations will generate standard operation procedure logs.
[0069] step 120, training an LLM-based operation diagnostic model with the historical operation logs and standard operation procedure logs;
[0070] The historical operation, historical operation logs and standard operation can be input to an LLM-based operation diagnostic model to train the operation diagnostic model. This training process enables the model to develop a deep understanding of machining tool operations. The LLM-based approach allows for nuanced interpretation of complex log data.
[0071] In some embodiments, training an LLM-based operation diagnostic model with the historical operation logs and standard operation procedure logs comprising: preprocessing the historical operation logs; utilizing the preprocessed historical operation logs to learn and generate sequence of logs by unsupervised learning; receiving labels from user for the sequence of logs and training the operation diagnostic model with the labels and the sequence of logs by supervised learning.
[0072] In detail, preprocessing the historical operation logs can include cleaning, normalizing, and potentially extracting features, to make them suitable for model training. The preprocessed historical operation logs are utilized for unsupervised learning. During this stage, a preliminary model is trained to learn the patterns and sequence in the logs. The goal of this model is to learn and generate the sequence of logs. Once the preliminary model has gained sufficient understanding of operational logic during the unsupervised learning stage, it is further fine-tuned through supervised learning. At this stage, generated logs are labeled with the type of operations by user and the labeled data is fed back to the model to learn the connection between logs and operations.
[0073] In some embodiments, utilizing the preprocessed historical operation logs to learn and generate sequence of logs by unsupervised learning comprising: utilizing the preprocessed historical operation logs, by a sequence-to-sequence model, to learn and generate sequence of logs by unsupervised learning. A sequence-to-sequence model is a type of neural network architecture used for transforming one sequence into another. It is particularly effective in tasks where the input and output are sequences of different lengths or where they may not align one-to-one. With use of sequence-to-sequence model, more complex data relationships and dependencies can be managed and performance in real-world applications can be improved.
[0074] step 130, obtaining current operation logs;
[0075] Current operation logs can be stored locally on the machine tool or centrally on the cloud. In one example, current operation logs can be obtained by accessing to the local machine tool. In other examples, current operation logs can be obtained by accessing to the cloud remotely.
[0076] step 140, diagnosing current operation corresponding to the current operation logs by the operation diagnostic model using the current operation logs.
[0077] The trained operation diagnostic model has capabilities to interpret current operation logs in the context of historical operation logs and standard operation procedure logs. The current operation logs are input into the operation diagnostic model. The operation diagnostic model can output current operations corresponding to the current operation logs and standard operations. By comparing the current operations and the standard operations, the operation diagnostic model can diagnose the current machine tool operations. If the current operations are in consistent with the standard operations, it shows current operation is correct. If the current operations are not in consistent with the standard operations, it shows current operation is not fully correct. The LLM's natural language processing abilities enable it to generate human-readable content, making the diagnostic output more accessible to operators.
[0078] Therefore, the present disclosure utilizes an LLM-based diagnostic model to automatically interpret and generate log entries, significantly reducing the need for specialized human expertise and speeding up the diagnostic process. Besides, the present disclosure offers real-time diagnostic feedback by comparing user-generated logs with standard operation procedures, immediately identifying discrepancies and suggesting corrective actions.
[0079] In some embodiments, the method further comprising: setting up a test set; generating an operation script, by the diagnostic model, with the test set; optimizing the diagnostic model with the test set and the operation script. In detail, a test set that includes machine tool operation logs that have not been seen before can be set up. Model's accuracy, robustness, and efficiency in generating operation scripts can be assessed. Based on the test results, adjustments and optimizations can be made to the model.
[0080] In some embodiments, the method further comprising: generating current operation corresponding to the current operation logs; calling an HMI playback interface to complete playback of the current operation; detecting operation error in the current operation; marking location and cause of the error operation based on standard operation. In detail, the operation diagnostic model refers to the user log operation sequence and calls the HMI playback interface to complete the playback of the user's current operations. When the HMI plays back the moment when the error occurs, the location and cause of the error are clearly marked, and based on the generated correct operation process.
[0081] In some embodiments, the method further comprising: detecting if error operation occurs in the current operation; providing a step-by-step guidance for user to correct the error operation. Furthermore, When the HMI plays back the moment when the error occurs, the location and cause of the error are clearly marked, and based on the generated correct operation process, the user is guided step by step to complete the remaining operations.
[0082] In some embodiments, user tells the operation diagnostic model the current intent of the operation. The operation diagnostic model will interpret the user’s intention and process accordingly. For example, if the user intends to diagnose the current operation, The operation diagnostic model will generate the standard operation procedure. Then, the operation diagnostic model can be feedbacked by the user if the output of the operation diagnostic model realizes the user’s intention.
[0083] Fig. 2 is an exemplary schematic diagram of a system for diagnosing machining tool operations according to an embodiment of the present disclosure.
[0084] As shown in Fig. 2, an LLM-based operation diagnostic model is trained with standard operation procedure logs 202 and historical operation logs 203 in the stage of training. Standard operation procedure logs 202 is generated from standard operations 201 based on manuals. The trained operation diagnostic model 23 has the capacity to interpret operation logs as operations. Besides, the operation diagnostic model 23 is based on LLM. The LLM's natural language processing abilities enable it to generate human-readable content, making the diagnostic output more accessible to operators.
[0085] In the stage of inference, current operation logs 204 is input into the operation diagnostic model 23. Current operation logs 204 is generated based on user 21 operation on machine tool 22. Fig. 2 shows three typical scenarios 24, 25 and 26. In scenario 24, the current operations are correct. In scenario 25, operation 3 is wrong and marked. In scenario 26, operation 3 is missed and marked.
[0086] Below is a specific example according to the present disclosure.
[0087] Current operations include 5 operations.
[0088] Operation 1: At “Machine” area, press softkey “Meas. Workp. ”
[0089] Operation 2: Choose “Rectang. spigot” as the measurement method.
[0090] Operation 3: Manually measure the length, width, and vertical distance from the workpiece to the probe and type the measured values to “L” , “W” , and “DZ” respectively.
[0091] Operation 4: Set “X0” and “Y0” .
[0092] Operation 5: Press “Cycle start” on MCP to start the measurement.
[0093] If at operation 2 a wrong method is selected, operation 2 will be marked and hinted by the operation diagnostic model. If at operation 3 and 4 the wrong dimension is measured and recorded, operation 3 and 4 will be marked and hinted by the operation diagnostic model.
[0094] Fig. 3 is an exemplary schematic diagram of a process for diagnosing machining tool operations according to an embodiment of the present disclosure. As shown in Fig. 3, the process comprises:
[0095] step 301, start;
[0096] step 302, the operation diagnostic model is informed by the user;
[0097] step 303, generating standard operation procedure logs by the operation diagnostic model;
[0098] step 304, comparing the standard operation procedure logs with the current operation logs, and confirming the steps and reasons for abnormal operation by the operation diagnostic model;
[0099] step 305, visualizing the current operation logs and location of the problem by the operation diagnostic model;
[0100] step 306, guiding the user to correct the error operation based on the generated standard operation procedures;
[0101] step 307, judging if it is confirmed by the user that the final result is satisfied? if yes, move to step 308, or move to step 302;
[0102] step 308, end;
[0103] The present disclosure provides an apparatus for diagnosing machining tool operations. Fig. 4 is an exemplary schematic diagram of an apparatus 400 for diagnosing machining tool operations according to an embodiment of the present disclosure. As shown in Fig. 4, apparatus 400 comprises:
[0104] a first obtaining module 410, configured to obtain historical operation, historical operation logs corresponding to the historical operation and standard operation;
[0105] a training module 420, configured to train an LLM-based operation diagnostic model with the historical operation, historical operation logs and standard operation;
[0106] a second obtaining module 430, configured to obtain current operation logs;
[0107] a diagnosing module 440, configured to diagnose the current operation by the LLM-based operation diagnostic model using the current operation logs.
[0108] Embodiments of the present disclosure also propose an electronic device with a processor memory architecture. Fig. 5 is an exemplary structural diagram of an electronic device according to an embodiment of the present disclosure. As shown in Figure 5, electronic device 500 includes a processor 510, a memory 520, and a computer program stored on memory 520 that can run on processor 510. When the computer program is executed by processor 510, the method for identifying building block as described in either of the above is implemented. Among them, memory 520 can be implemented as various storage media such as electrically erasable programmable read-only memory (EEPROM) , flash memory, programmable program read-only memory (PROM) , etc. Processor 510 can be implemented to include one or more central processors or one or more field programmable gate arrays, wherein the field programmable gate array integrates one or more central processor cores. Specifically, the central processing unit or core can be implemented as a CPU, MCU, DSP, and so on.
[0109] It should be noted that not all steps and modules in the above processes and structural diagrams are necessary, and some steps or modules can be ignored according to actual needs. The execution sequence of each step is not fixed and can be adjusted as needed. The division of each module is only for the convenience of describing the functional division used. In actual implementation, a module can be divided into 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 different devices.
[0110] The hardware modules in each implementation can be implemented mechanically or electronically. For example, a hardware module can include specially designed permanent circuits or logic devices (such as dedicated processors, such as FPGA or ASIC) to complete specific operations. Hardware modules can also include programmable logic devices or circuits temporarily configured by software (such as general-purpose processors or other programmable processors) for performing specific operations. As for the specific use of mechanical methods, either dedicated permanent circuits or temporarily configured circuits (such as software configuration) to implement hardware modules, it can be determined based on cost and time considerations.
[0111] The above is only a preferred embodiment of the present disclosure and is not intended to limit the scope of protection of the present disclosure. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.
[0112] Independent of the grammatical term usage, individuals with male, female or other gender identities are included within the term.
Claims
1.A method (100) for diagnosing machining tool operations, comprising:obtaining historical operation logs and standard operation procedure logs, the standard operation procedure logs is generated from standard operations based on manuals (110) ;training an LLM-based operation diagnostic model with the historical operation logs and standard operation procedure logs (120) ;obtaining current operation logs (130) ;diagnosing current operation corresponding to the current operation logs by the operation diagnostic model using the current operation logs (140) .2.The method (100) according to claim 1, training an LLM-based operation diagnostic model with the historical operation logs and standard operation procedure logs comprising:preprocessing the historical operation logs;utilizing the preprocessed historical operation logs to learn and generate sequence of logs by unsupervised learning;receiving labels from user for the sequence of logs and training the operation diagnostic model with the labels and the sequence of logs by supervised learning.3.The method (100) according to claim 2, utilizing the preprocessed historical operation logs to learn and generate sequence of logs by unsupervised learning comprising:utilizing the preprocessed historical operation logs, by a sequence-to-sequence model, to learn and generate sequence of logs by unsupervised learning.4.The method (100) according to any one of claim 1 to claim 3, the method (100) further comprising:setting up a test set;generating an operation script, by the operation diagnostic model, with the test set;optimizing the diagnostic model with the test set and the operation script.5.The method (100) according to claim 1, the method (100) further comprising:generating current operation corresponding to the current operation logs;calling an HMI playback interface to complete playback of the current operation;detecting operation error in the current operation;marking location and cause of the error operation based on standard operation.6.The method (100) according to claim 5, the method (100) further comprising:detecting operation error in the current operation;providing a step-by-step guidance for user to correct the error operation.7.An apparatus (400) for diagnosing machining tool operations, comprising:a first obtaining module, configured to obtain historical operation logs and standard operation procedure logs, the standard operation procedure logs is generated from standard operations based on manuals (410) ;a training module, configured to train an LLM-based operation diagnostic model with the historical operation logs and standard operation procedure logs (420) ;a second obtaining module, configured to obtain current operation logs (430) ;a diagnosing module, configured to diagnose current operation corresponding to the current operation logs by the operation diagnostic model using the current operation logs (440) .8.An electronic device, comprising a processor (510) and a memory (520) , wherein an application program executable by the processor (510) is stored in the memory (520) for causing the processor (510) to execute a method for identifying building block according to any one of claims 1-6.9.A computer-readable medium comprising computer-readable instructions stored thereon, wherein the computer-readable instructions for executing a method for identifying building block according to any one of claims 1-6.10.A computer program product comprising a computer program, upon the computer program is executed by a processor for executing a method for identifying building block according to any one of claims 1-6.