Motion Sensor Position Classification via Associative Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for monitoring human body positions in manufacturing settings are inefficient, requiring additional manpower, prone to human error, and can be invasive, making it difficult to accurately measure and track potentially harmful positions without generating unnecessary data.
Innovation Solution
A system using motion sensing input devices and associative memory classification to detect and classify body positions, allowing for the collection of metrics only for specific, predefined movements or positions of interest, thereby reducing data resources and maintaining a non-invasive monitoring process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If comprehensive motion data is collected and monitored, then measurement precision and reliability of safety monitoring is improved, but quantity of data generated increases and device complexity increases
Solution Approach 1:
The system extracts and monitors only specific body positions that pose safety risks, rather than collecting comprehensive motion data. The motion sensor and classification system are configured to identify and flag only hazardous postures (e.g., bending beyond safe angles, awkward positions), filtering out normal movements. This selective extraction reduces data volume while maintaining safety monitoring accuracy.
Solution Approach 2:
The monitoring system applies different levels of scrutiny to different body positions. Critical positions that exceed safety thresholds trigger detailed metric collection and alerts, while normal positions receive minimal or no monitoring. This localized quality approach ensures measurement precision for hazardous positions without generating unnecessary data for safe positions.
2Reliability
If continuous monitoring of all motions is implemented, then reliability of safety monitoring is improved, but loss of time and productivity decreases due to extensive data processing
Solution Approach 1:
The system pre-defines safety thresholds and hazardous position criteria before monitoring begins. Motion patterns and safety boundaries are established in advance, allowing the classification system to quickly compare real-time sensor data against predetermined safety zones. This preliminary configuration enables rapid decision-making without time-consuming analysis during monitoring operations.
Solution Approach 2:
The system performs partial monitoring by focusing only on critical safety parameters rather than analyzing every motion detail. When a potential hazard is detected, the system triggers detailed metric collection; otherwise, it uses simplified classification. This partial action approach maintains safety reliability while minimizing time spent on exhaustive data processing.
3Measurement precision
If detailed metrics are collected for all body positions, then measurement precision is improved, but device complexity and data management complexity increases
Solution Approach 1:
The monitoring system segments body monitoring into distinct functional modules: motion sensors capture raw data, a classification system identifies body positions, and a metric generation module collects detailed metrics only for hazardous positions. This segmentation allows each component to operate independently with optimized complexity, reducing overall system complexity while maintaining measurement precision for critical positions.
Solution Approach 2:
The patent introduces a classification system as an intermediary between motion sensors and metric collection. This intermediary layer filters and categorizes motion data, determining which positions require detailed metric generation. The classifier acts as a gatekeeper, reducing the burden on downstream systems and simplifying data management while preserving precision for safety-critical measurements.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method (1200) for identifying a motion of interest of an individual. The method includes collecting, at a computer (1304), motion sensor input data of motions of the individual from a motion sensor (1302) for an interval of time. The method further includes analyzing, using the computer (1304), the motion sensor input data using an analysis application (1308) having a set of classified predetermined motions of interest. The analysis application (1308) classifies a movement captured during the interval of time as a motion corresponding to one of a plurality of pre-determined motions of interest based on shared relative attributes. The method further includes generating an output providing notice of an identified predetermined motion of interest to a monitoring system (1412).