ML Process Control for Multi-Station Manufacturing Adjustment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing manufacturing systems face challenges in optimizing process parameters across multiple process stations due to batch fluctuations and tool wear, requiring frequent manual adjustments by experienced users, which can be inefficient and prone to errors.

Innovation Solution

A control device with two machine learning algorithms, one for initial optimization and another for continuous adjustment, connected through a combination module to form a global model, enabling decentralized or centralized control of manufacturing plants by refining process parameters and features, using machine learning to adapt to changes in material and tool conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If manual adjustments are performed by experienced users to optimize process parameters, then manufacturing quality can be maintained, but productivity decreases and the process becomes prone to human error

Engineering Contradiction:
Improvemanufacturing qualityVSAvoidproduction efficiency
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The control device automatically monitors process parameters and performs self-adjustments based on machine learning algorithms, eliminating the need for manual intervention by experienced users. The system serves itself by detecting process characteristics and autonomously optimizing parameters to maintain manufacturing quality while improving productivity.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical adjustment operations with an automated control system using machine learning algorithms. The control device substitutes human operators by processing data from detection devices and automatically adjusting process parameters, thereby eliminating human error and increasing production efficiency.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Manufacturing precision

If manual adjustments are performed frequently to account for batch variations and tool wear, then process optimization is achieved, but loss of time increases

Engineering Contradiction:
Improveprocess optimizationVSAvoidadjustment time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The control device continuously monitors process characteristics and continuously adjusts process parameters without interruption to production. The machine learning algorithm operates in real-time, eliminating the need for frequent manual stoppages and adjustments, thereby maintaining process optimization while minimizing time loss.

Inventive Principle:
Principle #20Continuity of useful action

Solution Approach 2:

The system implements a closed-loop feedback mechanism where detection devices continuously monitor process characteristics, feed this information to the control device, which then automatically adjusts process parameters. This continuous feedback loop ensures process optimization is maintained without requiring manual intervention or causing production delays.

Inventive Principle:
Principle #23Feedback

3Productivity

If automated control systems are implemented to reduce manual intervention, then productivity increases, but device complexity increases

Engineering Contradiction:
Improveproduction efficiencyVSAvoidcontrol system complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The control device is designed as a multi-functional system that combines detection capabilities, machine learning processing, and actuation functions in a single integrated unit. This universal device performs multiple tasks (monitoring, analyzing, adjusting) that would otherwise require separate systems, thereby increasing productivity while managing overall system complexity.

Inventive Principle:
Principle #6Universality (Multi-functionality)

4Adaptability or versatility

If machine learning algorithms are used for continuous optimization, then adaptability to material and tool changes improves, but device complexity increases

Engineering Contradiction:
Improveresponse to material and tool changesVSAvoidalgorithm complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The machine learning algorithm is pre-trained with knowledge about process parameters, material variations, and tool wear patterns before deployment. This preliminary training enables the system to quickly adapt to new conditions without requiring complex real-time calculations, thereby improving adaptability while managing computational complexity.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3891561B1Control device for controlling a manufacturing plant, manufacturing plant and method
Publication Date: 2024.09.11 ROBERT BOSCH GMBH
  • EP3891561B1 patent drawingFigure 1

AI summary

The invention relates to a control device for controlling a manufacturing plant (1), the manufacturing plant (1) comprising at least one process station (2a, b, c) for carrying out a manufacturing process, the manufacturing plant (1) and/or the process station (2a, b, c) having at least one process parameter (6) for open-loop and/or closed-loop control, the manufacturing plant (1) and/or the process station (2a, b, c) having at least one detection device for detecting at least one process feature (12). The control device has a first control module (7) and a second control module (11), the first control module (7) and the second control module (11) being designed to define a controlled value and/or a model for the process parameter (6) on the basis of a machine learning algorithm and the process feature (12) for closed-loop control of the manufacturing plant (1) and/or the process station (2a, b, c).