Action Recognition Certificates for Unbiased Process Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methodologies in manufacturing and other contexts face challenges in collecting comprehensive and unbiased data on human activities, leading to incomplete and biased insights, which hampers process optimization and quality improvement.
Innovation Solution
An action recognition and analytics system utilizing sensor streams from various sources, including video, thermal, and depth sensors, to automatically recognize cycles, processes, actions, and sequences, and create certificates for data sets, enabling more detailed and accurate analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual techniques are used to gather data on human activity, then the implementation is simple and low-cost, but the data set is small, incomplete, and biased
Solution Approach 1:
The patent replaces manual data collection methods with automated computer vision and machine learning systems. Sensors capture video and sensor data streams that are automatically processed by algorithms to detect actions, processes, and cycles, eliminating the need for manual observation and recording while providing comprehensive, unbiased data sets
Solution Approach 2:
The patent introduces computer vision algorithms and machine learning models as intermediaries between the physical human activity and the data representation. These intermediaries automatically interpret sensor streams, detect actions and processes, and generate structured data sets without direct human intervention in the measurement process
2Extent of automation
If IIoT devices are used to collect data, then automation is increased, but the data set remains incomplete as machines only perform a small portion of tasks
Solution Approach 1:
The patent creates a universal data collection system that can capture both machine-performed tasks and human-performed tasks within a single integrated framework. The system processes sensor streams from diverse sources including IIoT devices and video sensors to create comprehensive data sets that cover the entire manufacturing process regardless of whether the actor is human or machine
Solution Approach 2:
The patent segments the manufacturing process into discrete detectable units such as actions, processes, and cycles that can be independently identified and recorded. This segmentation allows the system to comprehensively capture both automated machine operations and manual human activities as distinct but integrated components of the overall process
3Productivity
If the number of actions per station increases, then productivity is improved, but the cognitive load on the operator increases resulting in higher deviation rates
Solution Approach 1:
The patent implements feedback mechanisms where the automated detection system continuously monitors operator actions and provides real-time or post-process validation. The system compares detected actions against expected processes, enabling immediate correction of deviations or identification of training needs, thereby maintaining high reliability even as productivity increases
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 automatic creation of certificates for each instance of a subject product or service. The certificate can string together snippets of the sensor streams along with indicators of cycles, processes, action, sequences, objects, parameters and the like captured in the sensor streams.