Search

JP-2026075605-A - Method and monitoring device for monitoring production machinery

JP2026075605AJP 2026075605 AJP2026075605 AJP 2026075605AJP-2026075605-A

Abstract

[Problem] To provide a method and monitoring device for monitoring production machinery with less effort. [Solution] Each quality parameter (Q1, Q2, ...) of each processed workpiece is acquired. Furthermore, if each operating signal sequence deviates from the target sequence, it is checked whether each operating signal sequence (B1, B2) is associated with a fault cause (F1, F2) in the monitoring database (DB). If this is not the case, each quality parameter (Q1) is output along with a request (REQ) for input of the fault cause, each fault cause (F1) is read, and each operating signal sequence (B1), each quality parameter (Q1), and the read fault cause (F1) are associated with each other and stored in the monitoring database (DB). In all other cases, the associated fault cause (F2) is output, and/or the production machine (PM) is controlled according to the associated fault cause (F2). [Selection Diagram] Figure 1

Inventors

  • ハラルト フェルクル
  • フアン・マヌエル ロレンツィ

Assignees

  • シーメンス アクチエンゲゼルシヤフト

Dates

Publication Date
20260508
Application Date
20251009
Priority Date
20241022

Claims (15)

  1. A computer-based method for monitoring production machinery (PM), wherein during the operation of the production machinery (PM), a) The current operating signal (BS) of the production machine (PM) is continuously acquired, b) For each of the multiple workpieces (W1, W2, ...), a time window (T1, T2, ...) is obtained in which each workpiece is processed by the production machine (PM), c) For each time window (T1, T2, ...), the progress of the operating signal (B1, B2, ...) limited to each of the said time windows is acquired. d) The quality parameters (Q1, Q2, ...) of each processed workpiece (W1, W2, ...) are obtained, e) As a result of detecting deviations from the target sequence (SV) for each operating signal sequence, it is checked whether a fault cause (F1, F2) is assigned to each of the aforementioned operating signal sequences (B1, B2) in the monitoring database (DB), and if this does not apply, - Information regarding each of the aforementioned quality parameters (Q1) is output via the user interface (IO) along with a request (REF) for input of each cause of failure. - Each of the aforementioned fault causes (F1) is read by the user interface (IO), - The respective operating signal progression (B1), the respective quality parameter (Q1), and the respective read fault cause (F1) are associated with each other and stored in the monitoring database (DB). Otherwise, - The associated fault cause (F2) is output, and/or the production machine (PM) is controlled according to the associated fault cause (F2). method.
  2. The method according to claim 1, Each of the aforementioned operating signal sequences (B1), each of the aforementioned quality parameters (Q1), and each of the read fault causes (F1) are associated with each other and stored in a knowledge graph (KG). A method characterized by the following.
  3. A method according to claim 1 or 2, The pattern recognition routine extracts characteristic signal patterns from the acquired operating signal sequences (B1, B2, ...). In order to detect the deviation from the target progress (SV), each of the extracted signal patterns is compared with the target signal pattern extracted from the target progress (SV) by the pattern recognition routine, and Each of the aforementioned operating signal sequences (B1, B2, ...) is stored in the monitoring database (DB) in the form of each extracted signal pattern. A method characterized by the following.
  4. A method according to any one of claims 1 to 3, The acquired operating signal sequences (B1, B2, ...) are subjected to cluster analysis by a clustering routine, and the acquired operating signal sequences (B1, B2, ...) are classified into clusters of similar operating signal sequences. In order to detect the deviation from the target progress (SV), it is checked whether each of the operating signal progresses (B1, B2, ...) is in the same cluster as the target progress (SV), and Each of the aforementioned operating signal sequences (B1, B2, ...) is stored in the monitoring database (DB) in the form of an identifier for the cluster to which it belongs. A method characterized by the following.
  5. A method according to any one of claims 1 to 4, If each of the aforementioned quality parameters (Q1, Q2, ...) does not meet the predetermined quality standard (QC), it is checked whether a cause of failure has been assigned to each of the aforementioned quality parameters in the monitoring database (DB), and If this is not the case, information regarding each of the aforementioned quality parameters, along with a request (REF) for input of each of the aforementioned failure causes, will be output by the user interface (IO). A method characterized by the following.
  6. A method according to any one of claims 1 to 5, The aforementioned target progress (SV) is derived from past operating signal progress in which the relevant quality parameters met predetermined quality standards (QC). A method characterized by the following.
  7. A method according to any one of claims 1 to 6, Each of the aforementioned quality parameters (Q1, Q2, ...) is determined using an optical sensor, a camera, and/or a sensor (S) for measuring geometric properties, electrical properties, mechanical properties, optical properties, or material properties, or for measuring surface roughness. A method characterized by the following.
  8. A method according to any one of claims 1 to 7, Along with the information relating to each of the aforementioned quality parameters, information relating to the progress of each of the aforementioned operating signals is also output by the user interface (IO). A method characterized by the following.
  9. A method according to any one of claims 1 to 8, The system checks whether a quality parameter has been assigned to each of the aforementioned operating signal sequences within the monitoring database (DB), and If this is the case, the assigned quality parameter is output, and/or the production machine (PM) is controlled according to the assigned quality parameter. A method characterized by the following.
  10. A method according to any one of claims 1 to 9, A large-scale language model (LLM) is provided that is pre-trained to generate relevant response text based on the input text. The stored operating signal history, quality parameters, fault causes, and their correspondences are supplied to the Large-Scale Language Model (LLM) in text format. The user interface (IO) reads the user query, inputs it into the large-scale language model (LLM) in the form of input text, and the large-scale language model (LLM) generates response text from it, and The response text is output by the user interface (IO). A method characterized by the following.
  11. A method according to any one of claims 1 to 10, If multiple fault causes are assigned to each of the aforementioned operating signal progressions within the monitoring database (DB), these fault causes are output by the user interface (IO) in order to select the appropriate fault cause. A method characterized by the following.
  12. A method according to any one of claims 1 to 11, The operation signal progress and/or quality parameters of multiple production machines are acquired and stored in a cross-machine monitoring database (DB), and, For the operation signal progression and/or quality parameters acquired in the first production machine, the cause of failure in a second production machine, which is different from that of the first production machine, is assigned within the monitoring database. A method characterized by the following.
  13. A monitoring device (MON) for monitoring production machinery (PM), configured to perform a method step of any one of claims 1 to 12.
  14. A computer program product comprising, during the execution of the program by a computer, an instruction causing the monitoring device (MON) described in claim 13 to execute the method described in any one of claims 1 to 12.
  15. A computer-readable storage medium storing the computer program product described in claim 14.

Description

Complex production processes typically require reliable quality monitoring of manufactured or processed products or workpieces. Ideally, the causes of quality defects should be identified as early as possible to effectively prevent future quality problems. Modern production machinery typically provides control functions that continuously evaluate current manufacturing data and, if predetermined tolerances are exceeded, for example, to halt the production process. However, in many cases, the root cause of quality problems cannot be identified or indicated using automated methods. Furthermore, many production machines cannot automatically learn to avoid this problem or similar problems in the future after an issue has occurred. In such cases, it is generally expected that operators manually identify the cause of the problem based on checklists and other methods and take appropriate corrective actions. However, such root cause identification is usually quite time-consuming. Figure 1 shows the monitoring of a production machine using a monitoring device according to the present invention.Figure 2 shows the progression of operating signals over time windows specific to multiple workpieces. Where identical or corresponding reference numerals are used in a figure, these reference numerals indicate identical or corresponding entities, which may be implemented or configured in particular as described in relation to the figure. Figure 1 shows the monitoring of a production machine PM by a monitoring device MON according to the present invention. The production machine PM may be a machine tool, robot, conveyor system, or manufacturing equipment, or may include such machines. The monitoring device MON comprises one or more processors PROC for performing method steps according to the present invention, and one or more memories MEM for storing data to be processed. The monitoring device MON is shown in Figure 1, positioned outside the production machine PM, and connected to it. Alternatively, the monitoring device MON may be incorporated, in whole or in part, within the production machine PM, or within the control unit for controlling the production machine. Preferably, the monitoring device MON is implemented, in whole or in part, within a so-called edge device. The production machine PM continuously processes workpieces W1, W2, ... . These workpieces W1, W2, ... may include finished products, partial products, intermediate products, or other workpieces to be processed. The machining of each workpiece W1 or W2 is performed within a time window T1 or T2, ..., and these time windows T1, T2, ... may overlap. The production machine PM acquires the corresponding time window T1 or T2, ... for each workpiece W1 or W2. The start of each time window is set to the start of machining of each workpiece W1 or W2, and the end of the time window is set to the end of this machining. Correspondingly, the time windows T1, T2, ... may each be represented by a pair of numerical values. Furthermore, the monitoring machine MON continuously acquires one or more current operating signals BS of the production machine PM. The operating signals BS can be extracted during the operation of the production machine PM and/or detected or measured by sensors. The operating signals BS are acquired as a time series over time. The operating signal BS may include, in particular, the current control signals, setpoints, control parameters, adjustment signals, measured values, sensor signals, environmental signals, monitoring signals, diagnostic signals, and/or error signals of the production machine PM. The operating signal BS can, for example, quantify the output, rotational speed, torque, travel speed, applied or acting force, temperature, pressure, available resources, resource consumption, hazardous substance emissions, wear, load or vibration, and/or the position, orientation, or machining state of the workpiece of the production machine PM. The operating signal BS, along with the acquired time windows T1, T2, ..., is continuously transmitted from the production machine PM to the monitoring device MON, where it is input to the selection device SEL of the monitoring device MON. The selection device is used, in particular, to select a time window-specific interval from each operating signal BS. For clarity of explanation, only the processing of a single operating signal BS is explicitly shown below. Furthermore, using the sensor system S, the machining quality of each workpiece W1 or W2, ... is continuously measured after machining or acquired by other means. For this purpose, the sensor system S may include, for example, a camera, an optical sensor, and/or sensors for measuring the geometric, electrical, mechanical, optical, and material properties of the workpiece W1 or W2, ..., or for measuring the surface roughness. The sensor system S may be at least partially integrated into the production machine PM or monitoring device MON, or mounted externally. T