Periodic Task Analysis Using Vision and Sensors for Progress Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to determine the completion percentage and progression rate of periodic tasks efficiently, necessitating improved monitoring and control in production processes.
Innovation Solution
A periodic task analysis system utilizing imaging devices and sensors to collect data, a task learning subsystem with CNN and transformer, and a task analyzing subsystem to generate real-time progress positions and completion percentages, providing guidance feedback to actors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual monitoring of periodic tasks is performed, then completion status can be tracked, but labor intensity and time consumption increase significantly
Solution Approach 1:
The patent replaces manual mechanical monitoring with an automated image processing system that uses cameras and computer vision algorithms to detect and analyze task completion status, eliminating the need for human observers to physically monitor each periodic task
Solution Approach 2:
The system creates visual copies of the work area through imaging devices and processes these copies through image analysis algorithms to determine completion status, allowing remote and automated monitoring without direct human intervention in the physical space
2Productivity
If automated image processing is implemented to determine completion percentage, then monitoring efficiency improves, but system complexity increases
Solution Approach 1:
The patent segments the complex monitoring task into distinct processing stages: image acquisition by cameras, pre-processing to enhance quality, feature extraction to identify relevant visual elements, and final analysis to determine completion percentage, making the overall system more manageable and implementable
Solution Approach 2:
The system introduces intermediate processing layers including image pre-processing modules and feature extraction algorithms that bridge the gap between raw camera images and final completion determination, simplifying the core decision-making process
3Reliability
If real-time progress tracking is implemented, then task optimization is enabled, but computational resources and processing power increase
Solution Approach 1:
The patent applies partial action by focusing image processing and analysis only on critical regions or key features of the periodic task rather than processing entire images at full resolution, reducing computational load while maintaining sufficient accuracy for completion determination
Solution Approach 2:
The system performs image analysis at periodic intervals synchronized with the periodic task being monitored, rather than continuously processing every frame, which reduces computational resource consumption while still providing real-time progress tracking capability
Data Source
AI summary
A periodic task analysis system which analyzes a task conducted by an actor periodically includes a data collecting subsystem including at least one of an imaging device configured to image a state where a plurality of fundamental work operations is performed periodically, and at least one sensing means collecting data from the state; a data receiving subsystem operatively coupled to the data collecting subsystem, where the data receiving subsystem is configured to receive at least one of a video file recorded by the imaging device and a sensor data recorded by the sensing means; a task learning subsystem operatively coupled to the data receiving subsystem, where the task learning subsystem is configured to identify one or more sub-tasks from at least one of the recorded video file and the recorded sensor data; and a task analyzing subsystem. A method for analyzing a periodic task is also provided.