Workcell Portal Monitoring for Predicted Human Entry Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing manufacturing systems lack effective methods to monitor and manage the interaction between humans and machinery in industrial environments, particularly in multi-cell workspaces, leading to potential safety hazards due to limited detection capabilities of 2D and 3D sensors, especially at entry points where humans may be partially occluded or unrecognized.
Innovation Solution
Implement a system where adjacent workcells share monitoring responsibilities by predicting human movement across portals using cameras, reducing computational load and redundancy by signaling expected entry or exit of humans between workcells, and ensuring machinery safety protocols are activated accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 2D and 3D sensors are used to monitor entry points in multi-cell workspaces, then detection capability is improved, but computational complexity and redundancy increase
Solution Approach 1:
The monitoring system is divided into multiple workcell-level monitoring systems, each responsible for a specific workcell. Each monitoring system independently monitors its own workcell entry points, eliminating the need for a single complex centralized system to process all entry point data from multiple workcells. This segmentation reduces computational complexity while maintaining detection capability at each local level.
Solution Approach 2:
The system predicts human movement across portals between workcells before actual entry occurs. By anticipating human movement patterns and pre-processing detection data, the system reduces redundant computational analysis when humans actually enter workcells, thereby lowering overall computational complexity while maintaining detection accuracy.
2Reliability
If redundant monitoring is implemented across all workcells, then safety coverage is improved, but productivity is reduced due to computational overhead
Solution Approach 1:
Safety monitoring responsibilities are segmented and distributed to individual workcell monitoring systems rather than being centralized. Each workcell maintains its own safety coverage through local monitoring, eliminating redundant computational processing across the entire facility while preserving comprehensive safety coverage at each workcell level.
Solution Approach 2:
Each workcell monitoring system independently performs its own safety monitoring and human detection functions without requiring centralized computational resources. This self-service approach allows each workcell to maintain full safety coverage autonomously, reducing overall computational overhead and improving productivity while preserving reliability.
3Adaptability or versatility
If centralized monitoring of all workcells is implemented, then coordination is improved, but device complexity increases
Solution Approach 1:
The monitoring architecture is segmented into independent workcell-level systems that each manage their own coordination locally. This eliminates the need for a complex centralized monitoring system while maintaining coordination capabilities within each workcell through distributed intelligence.
Solution Approach 2:
Multiple workcell monitoring systems are combined through standardized communication protocols and interfaces, allowing coordination between workcells without requiring a single complex centralized system. The merging of distributed simple systems achieves coordination functionality while avoiding the complexity of centralized control.
Data Source
AI summary
A monitoring system for monitoring a portal between a three-dimensional workcell and an adjacent three-dimensional workcells each of which includes controlled machinery, the portal being traversable by humans from either of the workcells into the adjacent workcell, the monitoring system including a plurality of cameras distributed throughout at least one of the workcell or the adjacent workcell and a controller for determining a proximity of a human between the adjacent workcells, where the controller is configured to electronically signal determined expected entry of the human into the adjacent workcell, and where the controller ignores detected features at the portal unassociated with humans except following receipt of the signal indicating expected entry of the human into the adjacent workcell through the portal.