Actuator Condition Estimation Using Readable Factory Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current control systems in factories lack an efficient method to estimate the condition of actuators, such as robots and NC processing devices, which hinders predictive maintenance and leads to potential equipment failures and downtime.
Innovation Solution
A control system comprising a factory system connected to a learning system via a network, where the learning system extracts records for machine learning to generate an estimation model for actuator conditions, enabling predictive maintenance by acquiring and constructing readable data from various nodes within the factory system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a learning system is introduced to generate estimation models for actuator conditions, then predictive maintenance capability is improved, but system complexity increases
Solution Approach 1:
A readable data construction device is introduced as an intermediary component between the factory system and the learning system. This device acquires data from multiple factory nodes, constructs readable data in a standardized format, and provides it to the learning system. This intermediary approach resolves the contradiction by adding predictive maintenance capability without requiring direct integration between the learning system and complex factory nodes, thus managing system complexity.
2Measurement precision
If data is acquired from multiple nodes in the factory system, then estimation accuracy is improved, but data processing complexity increases
Solution Approach 1:
The readable data construction device performs multiple functions: it acts as a data acquisition interface for multiple node types, a data formatter that creates standardized readable data, and a data preprocessing unit. By consolidating these multiple functions into a single universal device, the system achieves high estimation accuracy through multi-node data while avoiding the complexity of handling each node type separately.
3Measurement precision
If machine learning records are extracted and processed, then actuator condition estimation is improved, but processing time increases
Solution Approach 1:
The readable data construction device performs preliminary data processing by acquiring raw data from factory nodes and constructing standardized readable data before the learning system processes it for estimation. This preliminary action prepares the data in advance in an optimized format, reducing the processing time required by the learning system while maintaining high estimation accuracy.
Data Source
AI summary
A control system includes a factory system to control an actuator, and a learning system. The learning system extracts records for machine learning associated with the actuator from the factory system via a network, and generates an estimation model for estimating a condition of the actuator by machine learning using the records. The factory system includes a plurality of nodes including a control device to control the actuator, and a readable data construction device connected to the learning system via the network. The readable data construction device acquires data items associated with the actuator from at least one of the nodes other than the readable data construction device, and constructs readable data which is readable from the learning system. The readable data includes the data items associated with the actuator, and the learning system extracts the records from the readable data.