Field Device Predictive Monitoring for Compliance Breach Timing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for predictive monitoring of field devices struggle to accurately determine the remaining time until a monitored characteristic becomes non-compliant with specified requirements, especially when uncertainty and unpredictable time dependencies are involved, limiting the ability to prevent non-compliance.

Innovation Solution

A method that continuously monitors deviations between measured and reference values, applies filters to separate noise, and uses Monte Carlo simulations based on noise and deviation pairs to estimate the remaining time until compliance is breached, accounting for average rate of change and uncertainty without requiring fixed data rates or prior knowledge.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If Monte Carlo simulations are performed based on measurement errors from previous calibrations to determine next calibration time, then the prediction accuracy of compliance breach is improved, but additional information about calibration uncertainty is required which may not be available

Engineering Contradiction:
Improveprediction accuracyVSAvoidcalibration uncertainty information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The system uses the field device's own operational data and observed deviations to generate predictions, rather than relying on external calibration uncertainty information. The Monte Carlo simulation is fed with actual measured deviations from the field device during its operation, allowing the system to self-assess its compliance status without requiring additional calibration metadata

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The approach shifts from using calibration uncertainty parameters (which are static and often unavailable) to using operational deviation parameters (which are dynamically available during field device operation). By monitoring actual deviations between measured and reference values during operation, the system transforms the input parameters for prediction from calibration-era data to operation-era data

Inventive Principle:
Principle #35Parameter changes

2Productivity

If time intervals between service actions are extended to reduce costs and efforts, then productivity is improved, but the risk of operating non-compliant devices increases

Engineering Contradiction:
Improveservice action efficiencyVSAvoidcompliance assurance
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The system performs preliminary predictions of compliance breach timing before the actual breach occurs. By continuously monitoring deviations and running Monte Carlo simulations, the system forecasts when the field device will exceed compliance limits, allowing maintenance to be scheduled proactively rather than reactively. This enables extending service intervals while maintaining compliance assurance through advance warning

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system establishes a feedback loop where actual operational deviations are continuously measured, fed into the prediction model, and used to update remaining compliance time estimates. This real-time feedback mechanism allows dynamic adjustment of maintenance scheduling, enabling longer intervals when compliance is stable while catching degradation trends early

Inventive Principle:
Principle #23Feedback

3Measurement precision

If the frequency of determining deviations between measured and reference values is increased to improve prediction accuracy, then measurement precision is improved, but the data rate requirements and system complexity increase

Engineering Contradiction:
Improveprediction accuracyVSAvoiddata processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system applies filtering to the recorded deviations to separate noise from actual drift signals. Rather than using all raw deviation data equally, the filtering process selectively processes deviations to extract the meaningful compliance trend. This partial processing approach maintains prediction accuracy by focusing on relevant signal components while reducing the computational burden of processing every raw data point

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3739403B1Method of operating and of predictive monitoring of a field device
Publication Date: 2022.12.07 ENDRESSHAUSER GRP SERVICES AG
  • EP3739403B1 patent drawingFigure 1~2
  • EP3739403B1 patent drawingFigure 3~4
  • EP3739403B1 patent drawingFigure 5~6

AI summary

An improved method of operating a field device (1) for measuring and/or monitoring at least one measurement variable (p) and of predictive monitoring of a compliancy of at least one characteristic of said field device (1) to a requirement specified for said field device (1) is described, comprising the steps of: continuously monitoring said characteristic by: at consecutive times (ti) determining and recording a deviation (D(ti)) between a measured value (M(ti)) of a monitored variable (m) determined by said field device (1) and a reference value (R(ti)) of said monitored variable (m), wherein said deviations (D(ti)) are indicative of a degree of compliancy to said requirement, applying a filter to the recorded deviations (D(ti)),based on the deviations (D(ti)) and the filtered deviations (FD(ti)) determining a noise (N) superimposed on the filtered deviations (D(ti)), at the end of at least one monitoring time interval (MTI) determining a remaining time (RT) remaining until the deviations (D(ti)) will exceed said deviation range (DR) by: for at least two different deviation pairs (k), each comprising a first and a second deviation (D1k(t1k), D2k(t2k)) determined based on the filtered deviations (FD(ti)), determining a simulated value (SRTk) of the remaining time (RT) by performing a Monte Carlo simulation based on the noise (N) and the respective deviation pair (k), based on the simulated values (SRTk) determining the remaining time (RT), and generating an output informing about the remaining time (RT).