The Challenge Pilot Addresses

QUARK tackles the rigidity of current quality inspection in the plastics industry, where reprogramming and setup changes are slow and costly. It introduces a flexible, AI-based, human-centered inspection system capable of adapting dynamically to new products and defects.

Objectives

QUARK aims to integrate and validate JARVIS tools—specifically AR Interfaces and Object Recognition & Pose Estimation—in a real industrial environment at GERMANPLAST, a plastics manufacturing company. The main objectives are to reduce inspection setup and reconfiguration time, improve defect detection accuracy and improve operator acceptance through user-friendly AR-based interfaces. The system will be demonstrated on an existing UR10 collaborative robot station equipped with a multi-camera vision system, reaching TRL 7 validation by the end of the pilot.

v.2_JARVIS one pager template – Sprint 3Download

Key Technologies or Methods Being Developed or Tested

  • AI-based vision, synthetic data, AR interfaces, object recognition, human-in-the-loop feedback, and automated palletization.

QUARK – Pilot Project – Sprint #3