Sensor Assembly Quality Separation for Fault Source Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sensor arrangements in production lines require significant response time and manual intervention for changes, especially when faults occur, and cannot distinguish between tag defects and sensor impairments, leading to inefficiencies in quality control and process monitoring.
Innovation Solution
A sensor arrangement with matrix factoring algorithms that separates sensor quality values and resource quality scores, allowing for independent assessment and monitoring of sensor and resource qualities, enabling automatic control and maintenance decisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If code readers are used to record quality of reading processes, then reading quality can be monitored, but it cannot be determined whether incorrect reading is due to tag defect or sensor impairment
Solution Approach 1:
The patent segments the quality assessment into two independent components: sensor quality values and resource quality scores. This is achieved through matrix factorization that decomposes the quality matrix into sensor-specific and resource-specific factors, allowing independent evaluation of each component's contribution to reading quality or defects.
Solution Approach 2:
The patent introduces matrix factorization as an intermediary computational method that processes the quality data and separates the contributions of sensors and resources. This intermediary process enables the system to identify whether reading failures originate from sensor impairment or resource defects by analyzing the factorized components.
2Manufacturing precision
If manual intervention by planner is required for changes in production specifications, then control accuracy is maintained, but response time increases significantly
Solution Approach 1:
The patent enables the sensor arrangement to automatically monitor and assess its own quality parameters and resource conditions without requiring manual intervention. The system self-evaluates sensor performance and resource quality through automated data collection and matrix factorization, allowing rapid response to changes while maintaining control accuracy.
Solution Approach 2:
The patent implements automated feedback loops where quality data from sensors and resources is continuously collected, processed through matrix factorization, and used to generate actionable insights. This feedback mechanism enables the system to automatically detect and respond to quality issues or production changes without waiting for manual planner intervention.
3Reliability
If sensor quality and resource quality are assessed together, then overall system performance can be monitored, but specific fault sources cannot be identified
Solution Approach 1:
The patent applies segmentation by decomposing the overall quality assessment into sensor-specific quality values and resource-specific quality scores through matrix factorization. This mathematical decomposition separates the joint quality metric into independent components, making it possible to identify whether faults originate from sensors or resources while still monitoring overall system reliability.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
The invention relates to a sensor arrangement (100) with one or more sensors (201 - 204), with one or more operating means (101 - 104) each characterizing an object (801, 802) and with a processing unit (501) which is connected to the sensor(s) (201 - 204) via a communication link (401), wherein each sensor (201 - 204) is configured to read and uniquely identify an operating means (201 - 204). In each sensor (201 - 204), the sensor data includes the time of detection of an operating resource (101 - 104), the operating resource identification of the operating resource (101 - 104), a sensor identification uniquely identifying the sensor (201 - 204), and a time-, sensor-, and operating resource-dependent quality indicator q(t, s, b) determined for the detection of the operating resource (101 - 104) and transmitted to the processing unit (501).Using a matrix factoring algorithm implemented in the processing unit (501), sensor quality values q(t, s) and equipment quality values q(t, b) are separated from measured quality parameters q(t, s, b). The invention further relates to a corresponding method.