IL3.1C2 - Computer Vision System for Task Quality Control
An edge-computing computer-vision system for operator ergonomics and security in industrial tasks. It includes danger-zone detection, task monitoring and picking navigation, and hand/glove tracking, with laboratory validation under a Bosch-defined case study.
-
Main Gain(s): Operator safety monitoring, Task navigation support, Glove-aware assembly validation, Human digital-twin data capture
-
Core Contributions: Universidade de Aveiro (coord. Prof. Eugénio Rocha; support Diogo Costa) with Bosch Termotecnologia, S.A. (coord. Eng. Nelson Ferreira)
Use Cases:
-
Controlled laboratory case studies reproducing industrial picking tasks and Bosch safety scenarios involving moving bridge equipment
-
Start TRL: TRL 2
-
Final TRL: TRL 5
-
Main Features
- Danger-detection platform using regions of interest and intrusion simulation
- Task Monitoring and Picking Navigation with accuracy reported at or above 94.5% across tested scenarios
- Hand and glove tracking with multi-region validation
- Edge devices with GPU/TPU support and environmental sensing
- Laboratory setup replicating industrial picking and operator-safety scenarios