Machine Operating State Rules for Cross-Environment Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the normal operating state of machines struggle with adaptability to different jobs, environmental conditions, and variations in machine configurations, leading to inefficiencies and increased maintenance costs due to the need for individual adjustments and inability to account for changes in raw materials and environmental influences.
Innovation Solution
A method that establishes and adapts rules for characterizing the normal operating state of machines by using a standard machine to determine connections between measured values and environmental parameters, storing these rules in a central switching point, and applying them to similar or identical machines with adjustments for structural changes, sensor modifications, and environmental variations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If methods are individually adjusted for each machine and job configuration, then measurement precision and reliability are improved, but device complexity and time consumption increase significantly
Solution Approach 1:
The patent applies preliminary action by pre-determining characteristics of the normal operating state during a learning phase before actual production. The system collects measurement data during normal operation and automatically determines the characteristics without manual intervention, so that when production starts, the evaluation is already prepared and no time-consuming adjustments are needed.
Solution Approach 2:
The system applies self-service by automatically determining the characteristics of normal operating state without requiring manual configuration or adjustment by operators. The evaluation device autonomously collects data, processes it, and establishes the baseline for anomaly detection, eliminating the need for human experts to manually adjust parameters for each machine.
2Measurement precision
If comprehensive sensor coverage is implemented across all machine components, then measurement precision improves, but device complexity and cost increase
Solution Approach 1:
The patent applies the extraction principle by focusing measurement efforts only on the most relevant characteristics that define the normal operating state. Instead of monitoring all possible parameters from all sensors, the system identifies and extracts the key characteristics that are sufficient for anomaly detection, reducing the complexity of data processing while maintaining effective monitoring.
Solution Approach 2:
The system applies partial action by determining characteristics based on a representative subset of measurement data rather than processing all available sensor information. The learning phase uses a sufficient portion of normal operation data to establish characteristics, without requiring complete analysis of every measurement from every sensor, thus reducing computational complexity.
3Manufacturing precision
If methods are customized for specific machine configurations and jobs, then manufacturing precision improves, but adaptability to different machines and conditions deteriorates
Solution Approach 1:
The patent applies universality by creating a standardized evaluation device and method that can be used across different machine types and production tasks. The system determines characteristics of normal operating state in a unified manner that works for various machines, making the anomaly detection capability transferable and adaptable without requiring complete reconfiguration for each new application.
Solution Approach 2:
The system applies parameter changes by allowing the characteristics of normal operating state to be dynamically determined based on the specific machine and job conditions through the learning phase. Rather than using fixed thresholds, the system adapts its parameters to match the actual operating conditions, enabling both precision for the specific application and adaptability to different machines.
Data Source
Figure 1~2
AI summary
The invention relates to a method for detecting the normal operating state of a work process in which a machine (M') is operated in a work environment with external specifications, the machine (M') having sensors (S'1,..., S'm ) for measuring physical parameters at predetermined positions in the area of the machine M', with rules (R1,..., Rp) being searched for that relate the measured values determined during the normal operating state of the standard machine (M1) in the relevant working environment describe with the external specifications, and which are met during the normal operating state. The invention provides that the rules (R1,..., Rp) determined for the standard process are then applied to the working process of a machine similar or identical to the standard machine (M1). (M ') are transmitted in the work environment, it is examined whether the transmitted rules (R1, ..., Rp) by the measured values of those sensors (S'1, ..., S'm) of the machine e (M'), which correspond to the sensors (S1,..., Sn) of the standard machine (M1) in terms of arrangement and sensitivity, are fulfilled, and - if this is the case, the rules for characterizing the normal operating state of the working process are adopted, and - if this is not the case, the rules are adapted to the new measured values or discarded.