Production Video Analysis for Manual Process Deviation Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing manufacturing processes, particularly those involving manual tasks, lack effective condition monitoring, leading to undetected defects and increased waste due to inadequate inspection of production steps.
Innovation Solution
A method that analyzes a live video of a production process by comparing it to a reference video, identifying discrete tasks in both, and outputting an alert for deviations, using object detection algorithms and neural networks to ensure accurate task execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If condition monitoring is used for automated production steps, then manufacturing precision is improved, but it has minimal role for manual tasks
Solution Approach 1:
The video analysis system serves multiple functions: it monitors both automated machine-based production steps and manual tasks (folding, packing, wrapping) using the same technological approach. The system can identify tasks performed by both machines and human operators, making the condition monitoring capability universal across different production types.
Solution Approach 2:
The patent replaces traditional mechanical condition monitoring sensors with a vision-based system using cameras and machine learning algorithms. This substitution enables monitoring of manual tasks that are difficult to instrument with traditional sensors, while maintaining the ability to monitor automated equipment.
2Reliability
If traditional inspection methods are used, then device complexity is low, but defects are undetected leading to increased waste
Solution Approach 1:
The system creates a digital copy of the correct production process through a reference video, which is then broken down into discrete task frames. This reference model is stored and used for comparison against live production videos, enabling automated defect detection without complex physical inspection equipment.
Solution Approach 2:
The system continuously compares live video frames against the reference video frames and provides immediate feedback when deviations are detected. The comparison algorithm identifies mismatches between expected tasks (from reference) and actual tasks (from live video), enabling real-time defect detection.
3Productivity
If video analysis with frame comparison is implemented, then productivity is improved through real-time detection, but device complexity increases
Solution Approach 1:
The video analysis system segments the production process into discrete, non-overlapping task frames. Each frame represents a specific task (e.g., folding, packing) and is compared independently against the reference. This segmentation simplifies the comparison process and enables efficient real-time analysis.
Solution Approach 2:
The reference video is pre-processed to identify and store discrete task frames before actual production monitoring begins. This preliminary action creates a ready-to-use reference model that speeds up real-time comparison operations, reducing computational complexity during live monitoring.
Data Source
AI summary
A computer-implemented method, according to one embodiment, includes identifying, in a reference video of a production process of a product, a discrete and non-overlapping set of first tasks. The first tasks define at least a first sub-process. The method further includes analyzing a live video of the production process to identify frames of the live video that include second tasks that define a second sub-process, and analyzing the frames of the live video for determining whether a match exists between the first tasks and the second tasks. In response to a determination that the match does not exist, an alert that a deviation is present in the production process depicted in the live video is output.