Worker Task Analysis Using Position and Work Tendency Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing work analysis systems struggle to accurately discriminate between multiple works performed by workers in a workplace, particularly in areas where multiple tasks overlap, relying solely on image recognition is insufficient.
Innovation Solution
A work analysis device that utilizes image recognition combined with work tendency information, such as assignment and standard work periods, to probabilistically estimate the tasks performed by workers, adjusting probabilities based on positional data and temporal relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If workers perform multiple tasks in shared work areas, then work flexibility and productivity are improved, but it becomes difficult to accurately discriminate and estimate the specific tasks being performed
Solution Approach 1:
The system uses feedback loops where the controller continuously refines probability estimates by comparing calculated work probabilities with actual work tendency information. The controller adjusts probability values based on feedback from work tendency data, enabling accurate task discrimination even when workers move between multiple work areas performing different tasks.
2Device complexity
If the system uses only position information to determine work content, then the system complexity is reduced, but the accuracy of work estimation decreases when tasks are similar or workers move between areas
Solution Approach 1:
The system merges multiple information sources including position information, work area definitions, and work tendency information into a unified probability calculation framework. The controller combines these diverse data types to compute work probabilities, achieving high estimation accuracy without requiring complex additional sensing infrastructure.
3Measurement precision
If the system calculates work probability based on multiple factors including work tendency information, then work estimation accuracy is improved, but the computational complexity and processing time increase
Solution Approach 1:
The system performs preliminary actions by pre-defining work areas and storing work tendency information before actual work analysis begins. This pre-processing allows the controller to efficiently calculate probabilities during operation without performing complex computations in real-time, reducing processing complexity while maintaining high accuracy.
Data Source
AI summary
A work analysis device is for generating information on works performed by at least one worker in a workplace, and includes: a storage configured to store map information indicating an area associated with each of the works; an acquisition circuitry configured to acquire position information including positions of the at least one worker at each time in the workplace; and a controller configured to: (ii) perform arithmetic processing to discriminate the works at each time based on the position information and the map information; (ii) generate probability information indicating a probability that a work associated with an area including positions of the at least one worker at each time in the map information is performed by the at least one worker based on the position information; and (iii) adjust the probability information according to work tendency information indicating a tendency for the at least one worker to perform the works.