Edge-Executed Skills for Adaptive Industrial Robotics (EDGE-SKILLS)

Purpose and goal

The purpose of the project is to enable safe and predictable AI‑based robot programming for complex grasping and handling tasks in industrial environments. The goal is to combine advanced offline training with robust execution on edge hardware, increasing flexibility, reducing deployment time, and strengthening competitiveness, particularly for small and medium‑sized enterprises, while enabling more resource‑efficient production.

Expected effects and result

The project is expected to deliver a validated framework for AI‑based robot programming of complex grasping tasks on industrial edge hardware. Results include optimized models, established workflows, and verified performance in production environments. The effects are increased flexibility, reduced deployment time, and more resource‑efficient production through reduced waste, rework, and inefficiencies.

Planned approach and implementation

The project is carried out in collaboration between academia and industry through six work packages covering requirements analysis, data collection, model development, edge optimization, and deployment and validation in production environments. AI models are trained offline and adapted for safe, predictable execution. Solutions are iteratively tested in real robotic cells to ensure robustness, usability, and industrial integration.