Maintenance Learning Model for Pre-Visit Faulty Part Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional diagnostic techniques for air conditioners cannot reduce the number of site visits by maintenance operators, even after identifying abnormal causes or anomaly locations, as they require on-site inspection to identify faulty parts.
Innovation Solution
An apparatus and method that acquire device and operation information, along with event data, to perform learning and identify parts or procedures needed for maintenance, allowing preparation before site visits, thereby reducing the number of on-site services.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional diagnostic techniques are used to identify abnormal causes and anomaly locations, then diagnostic accuracy is improved, but the number of site visits by maintenance operators cannot be reduced
Solution Approach 1:
The system performs preliminary identification of faulty parts and maintenance procedures before the maintenance operator arrives at the site. By analyzing operation information and event data in advance, the system predicts anomaly locations and prepares maintenance plans, allowing operators to prepare necessary parts and procedures beforehand, thereby reducing the number of site visits needed
Solution Approach 2:
A maintenance management server acts as an intermediary between the air conditioner system and maintenance operators. The server collects operation information from multiple air conditioners, performs centralized analysis to identify faulty parts, and transmits maintenance information to operators, eliminating the need for operators to perform preliminary diagnostic inspections at each site
2Measurement precision
If maintenance operators perform on-site inspection to identify faulty parts, then identification accuracy is improved, but maintenance efficiency deteriorates due to repeated site visits
Solution Approach 1:
The system replaces the mechanical inspection process performed by maintenance operators with an automated information processing system. The maintenance management server analyzes operation information and event data to automatically identify faulty parts and determine maintenance procedures, substituting human physical inspection with automated diagnostic algorithms
Solution Approach 2:
The system establishes a feedback loop where operation information from air conditioners is continuously collected, analyzed by the maintenance management server, and used to generate maintenance information that is transmitted back to operators. This feedback mechanism enables continuous improvement of diagnostic accuracy without requiring repeated site visits
Data Source
AI summary
An apparatus for assisting maintenance work includes circuitry configured to perform learning based on a data set in association with a replaced or repaired part, or a new part after replacement indicated by work content information, and the learning includes processing an input of the data set or a portion of the data set, in accordance with model parameters of a machine learning model; determining work content information applied to the input; and updating the model parameters of the machine learning model based on the determined work content information.