Equipment Malfunction Diagnosis Using Multi-Model Validation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for diagnosing equipment malfunctions rely on manual troubleshooting, which is time-consuming and prone to human error, and require user knowledge to instruct automated systems effectively.

Innovation Solution

An electronic diagnostic device employs multiple machine-learning models to autonomously diagnose equipment malfunctions by acquiring input data, generating hypotheses, retrieving historical data, determining valid hypotheses, and displaying them for operator action, without user knowledge dependency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual troubleshooting by skilled technicians is used, then the diagnosis can be performed with existing knowledge, but the process is time-consuming and costly

Engineering Contradiction:
Improvediagnosis accuracyVSAvoidtroubleshooting time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs self-diagnosis by automatically analyzing sensor data and generating hypotheses about malfunctions without requiring manual intervention from technicians. The electronic diagnostic device independently processes input data, applies machine learning models, and produces diagnostic results, enabling the system to serve itself in the diagnosis process.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical troubleshooting processes with an automated electronic diagnostic system. The electronic diagnostic device uses machine learning models and algorithms to substitute the human technician's analytical work, transforming the mechanical/manual diagnosis process into an automated computational system that processes data and generates hypotheses.

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

2Loss of time

If automated diagnostic solutions are implemented, then troubleshooting time is reduced, but the system requires user knowledge to instruct the model effectively

Engineering Contradiction:
Improvediagnosis timeVSAvoiduser knowledge requirement
Core Design Contradiction:
Loss of timeVSEase of operation

Solution Approach 1:

The electronic diagnostic device autonomously performs the complete diagnostic process without requiring user instructions or knowledge. The system independently acquires sensor data, applies multiple machine learning models to generate and validate hypotheses, and produces diagnostic results. The self-service capability eliminates the need for users to understand or direct the diagnostic process.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system transforms the diagnostic approach by changing the operational parameters from user-directed to autonomous operation. The machine learning models are configured to automatically process sensor data and generate hypotheses without user intervention, fundamentally altering how the diagnostic system operates and removing the dependency on user knowledge.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If multiple machine learning models are applied to generate and validate hypotheses, then diagnosis precision is improved, but device complexity increases

Engineering Contradiction:
Improvediagnosis precisionVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The diagnostic system is segmented into distinct functional modules: a first machine learning model for generating hypotheses, a second model for retrieving historical data, and a third model for validating hypotheses. Each model performs a specific function in the diagnostic process, allowing the complex diagnostic task to be divided into manageable, specialized components that work together systematically.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary validation mechanism where the third machine learning model acts as a mediator between hypothesis generation and final diagnosis. This intermediary model validates the hypotheses generated by the first model using historical data from the second model, ensuring diagnosis precision while maintaining a structured, modular system architecture.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP4708051A1Method for diagnosing a malfunction of a piece of equipment
Publication Date: 2026.03.11 THALES SA
  • EP4708051A1 patent drawingFigure 1
  • EP4708051A1 patent drawingFigure 2
  • EP4708051A1 patent drawingFigure 3

AI summary

Method for diagnosing a malfunction of an equipment, the method being implemented by an electronic device and comprising: - acquisition (100) of input data comprising information relating to the malfunction of the equipment, - applying (200) a first model to the input data to obtain an hypothesis including a solution of the malfunction, said first model being trained to determine the hypothesis about the malfunction based on such input data, - applying (300) a second model to the hypothesis to obtain a plurality of retrieval facts from a knowledge base containing historical data about the malfunction of the equipment, - determining (400) a valid hypothesis by applying a third model to the hypothesis and the retrieval facts, said third model being trained to determine whether a hypothesis is valid, - sending (500) a display control for displaying the valid hypothesis for diagnosis of the malfunction of the equipment.