Audi tests humanoids for parts sequencing

Audi and Agile Robots are testing mobile humanoids to pick components and arrange them in assembly order at Audi’s Production Lab. Announced on 21 October with Dassault Systèmes and NVIDIA, the proof of concept targets parts sequencing for Audi’s Ingolstadt plant.[1]

The robots combine a mobile base, two arms and AI perception to handle varied components and positions inside containers. The partners are creating a virtual twin of the logistics environment before deployment in the plant. The announcement gives no trial cycle times or picking success rates.[1]

A successful grasp can still cause a queue

Agile Robots’ separate technical account explains why the simulation extends beyond the gripper. Engineers model racks, components and the robot together: shifting a rack by centimetres can change the approach, while missing detail can hide a grasping problem.[2]

The model then expands to material flows and other machines. A robot may finish its handling task successfully yet create a bottleneck because of its timing. Comparing those interactions helps engineers decide which layouts and operating sequences deserve physical testing.[2]

Dassault’s robotics documentation describes another part of that preparation: virtual commissioning tests communication between robot controllers, industrial control systems and other automation. Checks include signals and timing. For a sequencing task, that means examining coordination with surrounding equipment as well as whether the robot can reach and move a part.[3]

Report an error or suggest a correction