Dynamic Incident Rate Diagnosis for Printer Component Maintenance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for diagnosing image forming devices, such as printers, rely on Mean Time Between Failure (MTBF) calculations, which are inadequate for devices operating under varying workloads and conditions, leading to inefficient identification of degraded components.
Innovation Solution
A method using a logistic regression model to monitor incident counts over time intervals, calculating error ratios, and setting threshold levels to determine when electro-mechanical parts require repair or servicing, allowing for differential diagnosis based on specific conditions and workloads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If MTBF calculations are used to diagnose image forming devices, then a general failure prediction method is provided, but the diagnostic accuracy deteriorates under varying workloads and conditions
Solution Approach 1:
The patent changes the diagnostic parameters from static MTBF values to dynamic incident counts that adapt to varying workloads and operational conditions. By monitoring incident rates relative to actual device usage and environmental factors, the system achieves both adaptability to different workloads and maintained diagnostic accuracy through context-aware threshold adjustments.
2Device complexity
If linear regression functions are used to calculate failure probability, then a simple prediction model is provided, but the model accuracy deteriorates when devices operate under different customer conditions
Solution Approach 1:
The patent transitions from static linear regression models to dynamic monitoring systems that continuously track incident counts and adjust diagnostic thresholds based on actual operational conditions. This dynamic approach maintains model simplicity while improving accuracy by adapting to different customer conditions and workload patterns without requiring complex calculations.
3Ease of operation
If average incident rates are used to classify problematic devices, then a standardized classification method is provided, but the classification accuracy deteriorates for devices with varying usage patterns
Solution Approach 1:
The patent applies local quality by customizing diagnostic thresholds and incident rate criteria according to specific device usage patterns, operational conditions, and device types. Instead of a single average threshold for all devices, the system adjusts parameters locally to match each device's operational context, thereby maintaining classification simplicity while improving identification accuracy for devices with varying usage patterns.
Data Source
AI summary
A method to diagnosis an electro-mechanical part on an image forming device comprising: determining an incident count for a first time interval for the electro-mechanical part; determining a first threshold level for the first time interval; monitoring the electro-mechanical part for a second time interval, wherein the second time interval is a plurality of first time intervals; determining if the incident count for a first portion of the second time interval for the electro-mechanical part equal to or less than the first threshold; determining if the incident count for a second portion of the second time interval for the electro-mechanical part is equal to or less than the first threshold when the incident count for the first portion of the second time interval fails to exceed the first threshold for the electro-mechanical part; determining a maximum incident count for the second portion of the second time interval when the incident count for the second portion of the second time interval exceeds the first threshold for the electro-mechanical part; calculating an error ratio between an average incident count of the first portion of the second time interval and the maximum incident count of the second portion of the second time interval; and marking the electro-mechanical part as needing to be one of repaired or serviced when the error ratio exceeds an error ratio threshold level.