Sensor Maintenance Timing from Time-Series Data and ML Labels

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current sensor maintenance practices in industrial plants are inefficient due to varying deterioration rates and environments, leading to increased maintenance costs and difficulty in determining the correct maintenance time for sensors, as existing methods rely heavily on operator experience and regular, often unnecessary, maintenance schedules.

Innovation Solution

An information processing device and method that utilizes machine learning to analyze time-series measured data from sensors, generating a determination model to predict maintenance times and reduce unnecessary maintenance by identifying the need for zero-point adjustment, cleaning, or replacement based on data analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If regular maintenance schedules are implemented for sensors, then sensor reliability is maintained, but maintenance costs increase due to unnecessary maintenance on sensors that do not require adjustment and replacement

Engineering Contradiction:
Improvesensor reliabilityVSAvoidmaintenance costs
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The patent changes the maintenance approach from fixed time-based scheduling to condition-based scheduling by monitoring sensor parameters (measured values, standard deviations, drift rates) to determine when maintenance is actually needed, thereby avoiding unnecessary maintenance while ensuring reliability

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent replaces manual operator judgment and regular mechanical maintenance schedules with an automated information processing system that uses machine learning models to predict sensor deterioration and determine optimal maintenance timing

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

2Measurement precision

If machine learning is applied to sensor data analysis, then maintenance timing accuracy is improved, but device complexity increases due to the need for determination models and data processing systems

Engineering Contradiction:
Improvemaintenance timing accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs self-learning by automatically acquiring sensor data, processing it through determination models, and improving its predictive accuracy over time without requiring external intervention, thereby managing complexity through automation

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent introduces an information processing device as an intermediary between sensors and maintenance decisions, which handles the complexity of data analysis and model application, leaving the actual sensor operation simple and unchanged

Inventive Principle:
Principle #24Intermediary (Mediator)

3Ease of operation

If operator experience is used to determine maintenance time, then flexibility is maintained, but measurement precision deteriorates due to subjectivity and variability in operator judgment

Engineering Contradiction:
Improveoperational flexibilityVSAvoidmaintenance timing accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The system incorporates feedback loops where sensor data is continuously monitored, compared against determination models, and used to refine predictions of maintenance timing, objectively capturing and applying operational knowledge that was previously subjective

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3321756B1Information processing device, information processing method, program, and storage medium
Publication Date: 2021.01.27 YOKOGAWA ELECTRIC CORP
  • EP3321756B1 patent drawingFigure 1
  • EP3321756B1 patent drawingFigure 2
  • EP3321756B1 patent drawingFigure 3

AI summary

An information processing device according to one aspect of the present invention includes a first acquirer configured to acquire measured data of a sensor, a second acquirer configured to acquire maintenance information related to maintenance performed on the sensor, a learner configured to learn teacher data in which the acquired maintenance information as label information is associated with the acquired measured data to generate a determination model, and a storage storing the generated determination model.