Knitting Workflow Action Recognition for Real-Time Packing Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current manufacturing processes face challenges in collecting comprehensive and unbiased data on human activities, leading to incomplete and biased insights, which hampers quality improvement and process optimization in industries like manufacturing, healthcare, and retailing.
Innovation Solution
An action recognition and analytics system utilizing deep learning and sensor streams from various sources, including video, thermal, and depth sensors, to recognize cycles, processes, and actions, providing real-time verification and data for improved process management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual techniques are used to gather data on human activity, then data collection is simple to implement, but the data set is small, incomplete, and biased
Solution Approach 1:
The system employs multiple sensor types (video cameras, depth sensors, thermal sensors, RFID readers, force sensors, accelerometers, gyroscopes) that can simultaneously capture various aspects of human activity and machine interactions, creating a comprehensive and unbiased data set through multi-functional data collection
Solution Approach 2:
The patent introduces automated sensing systems as intermediaries between human operators and the data collection process, eliminating the need for manual observation and recording while providing continuous, objective measurement of cycles, processes, actions, and sequences
2Extent of automation
If IIoT devices are used to collect data, then automation is increased, but data collection is limited to only machine tasks
Solution Approach 1:
The system merges machine sensor data with human activity sensor data into a unified data set, combining IIoT device capabilities with specialized human motion tracking sensors to capture both machine operations and human actions simultaneously
Solution Approach 2:
The sensing system is designed to universally capture data from both machines and human operators using the same infrastructure, ensuring that no information is lost and providing a complete view of the manufacturing process
3Productivity
If the number of actions per station increases, then productivity is improved, but cognitive load on operators increases
Solution Approach 1:
The system provides real-time feedback to operators through the sensing and analysis infrastructure, monitoring their actions and providing guidance to reduce cognitive load while maintaining high productivity levels through automated tracking and performance optimization
4Manufacturing precision
If process optimization is pursued, then quality improvement is achieved, but data collection delays occur
Solution Approach 1:
The system implements continuous data collection and real-time analysis through always-on sensors and streaming data processing, eliminating delays between data collection and analysis while maintaining continuous process optimization and quality improvement
Data Source
AI summary
The systems and methods provide an action recognition and analytics tool for use in manufacturing, health care services, shipping, retailing and other similar contexts. Machine learning action recognition can be utilized to determine cycles, processes, actions, sequences, objects and or the like in one or more sensor streams. The sensor streams can include, but are not limited to, one or more video sensor frames, thermal sensor frames, infrared sensor frames, and or three-dimensional depth frames. The analytics tool can provide for kitting products, including real time verification of packing or unpacking by action and image recognition.