Abnormality Detection Using Operator Voice Data and Supervised Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional abnormality detection in production apparatuses relies solely on sensors, which may not capture sounds indicative of abnormalities from the surroundings, limiting the effectiveness of detection.
Innovation Solution
A system that includes a learning model construction device to acquire and analyze voice data from operators near production apparatuses, using supervised learning to construct a model for abnormality detection, incorporating features like speech content, tone, and volume, and integrating with sensors for comprehensive abnormality assessment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only sensor data from production apparatus is used for abnormality detection, then the detection system is simple and easy to implement, but the detection accuracy and coverage are limited because sounds from surroundings and operator reactions are not captured
Solution Approach 1:
The patent merges multiple data sources (sensor data from production apparatus, voice data from operators, and surrounding environment sounds) into a unified abnormality detection system. This combination allows the system to capture comprehensive information including operator reactions and environmental cues that sensors alone cannot detect, thereby improving detection accuracy while justifying the increased system complexity through enhanced functionality.
Solution Approach 2:
The detection system is designed to perform multiple functions: it not only detects mechanical abnormalities through sensors but also captures operator verbal reactions, analyzes surrounding sounds, and integrates these diverse data types for comprehensive abnormality assessment. This multi-functional approach enables the system to detect both equipment failures and operational anomalies, improving overall detection precision.
2Reliability
If voice data from operators is collected and analyzed using supervised learning, then the abnormality detection capability is enhanced by capturing operator reactions and surrounding sounds, but the device complexity and data processing requirements increase
Solution Approach 1:
The system performs preliminary actions by collecting and storing voice data from operators during normal operations before actual abnormality detection is needed. Supervised learning models are trained in advance using this pre-collected data, creating ready-to-use detection capabilities. This preliminary data collection and model training phase enables the system to quickly and reliably detect abnormalities when they occur, improving reliability while managing complexity through advance preparation.
Solution Approach 2:
The system employs supervised learning models that automatically analyze voice data and identify patterns indicative of abnormalities without requiring constant human intervention. The models self-adjust and improve by learning from training data, enabling automated abnormality detection that enhances reliability while reducing the need for complex manual analysis systems.
Data Source
AI summary
To provide a learning model construction device, abnormality detection device, abnormality detection system and server for performing abnormality detection using sound information of the surroundings of a production apparatus. A learning model construction device includes a voice acquisition unit that acquires voice data including the voice of an operator located in the vicinity of a production apparatus, via a mic; a label acquisition unit that acquires an abnormality degree related to a production line including the production apparatus as a label; and a learning unit that constructs a learning model for the abnormality degree, by performing supervised learning with a group of voice data and label as training data.