Work Skill Support Using Skeleton-Based Non-Standard Task Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing systems for supporting worker skills in production environments struggle to efficiently identify and address low-skill work performance, as determining inefficient work through moving image analysis is complex and cumbersome.

Innovation Solution

A work skill supporting device that includes a storage unit for non-standard work models, a time series skeleton information acquisition unit, a non-standard work extraction unit, a part specification unit, and a work content specification unit, which analyze workplace images to identify and specify non-standard work, enabling targeted skill improvement support.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If moving image analysis is used to determine inefficient work, then worker skill level can be identified, but the determination process becomes complicated requiring adjustment for each work unit

Engineering Contradiction:
Improveskill level identification accuracyVSAvoiddetermination process complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts and analyzes specific skeletal features (hand positions, body posture, movement trajectories) from workplace images to determine non-standard work. By focusing on extracting only the necessary skeletal information rather than analyzing entire moving images, the system achieves accurate skill level identification while simplifying the determination process and eliminating the need for complex adjustments for each work unit.

Inventive Principle:
Principle #2Taking out (Extraction)

2Measurement precision

If moving image generation is performed for each work unit to determine inefficiency, then accurate identification is possible, but the process becomes cumbersome and complex

Engineering Contradiction:
Improvework efficiency identification accuracyVSAvoidoperational simplicity
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The system automatically captures workplace images, extracts skeletal information, and determines non-standard work without requiring manual generation or processing of moving images for each work unit. The automated skeletal analysis performs the determination directly from captured images, eliminating the cumbersome process of creating and analyzing separate moving images while maintaining accurate identification of work efficiency issues.

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS11176360B2Work skill supporting device and work skill supporting system
Publication Date: 2021.11.16 HITACHI LTD
  • US11176360B2 patent drawing
  • US11176360B2 patent drawing
  • US11176360B2 patent drawing

AI summary

A work skill supporting device include a storage unit that stores non-standard work model information including a condition of non-standard work, work procedure information including information indicating a work content and information indicating a part to be used in work, and a workplace internal image obtained by photographing an inside of a workplace; a time series skeleton information acquisition unit that acquires time series skeleton information of one or a plurality of workers from the workplace internal image; a non-standard work extraction unit that determines whether or not the time series skeleton information satisfies the condition; a part specification unit that specifies a part serving as a work target using the workplace internal image for the non-standard work determined as satisfying the condition; and a work content specification unit that specifies a work content of the non-standard work with reference to the work procedure information.