R10-D10: Helper of the Future
Between warehouse shelves, a small robot buzzes through the production hall. It grabs, transports, avoids obstacles, efficiently and unspectacularly, and it understands human commands to support workers. That is precisely the idea. R10-D10 is an autonomous helper that quietly does its job—and in doing so fulfills a central promise of AI.
The robot, a mobile manipulator initially created by Robotnik, is being further developed at THI's AI mobility hub AImotion Bavaria. Its concept is based on a real need: automating routine activities without complex technology, systems that are difficult to operate or costly maintenance. "R10-D10 is not a show robot. We build systems that really solve problems—and not just impress," emphasizes Alexander Schiendorfer, research professor for AI-based optimization in automotive production. The aim is for R10-D10 to be easily programmed by employees themselves and to be able to continue learning from experience, using a stack of AI-technologies including reinforcement learning, large language models, and knowledge graphs.
Intelligent Support in Warehousing and Production
"It is a kind of an 'automated helper’ that supports where there is a lack of workforce," says Schiendorfer. Ideally, it takes over the retrieval of workpieces from the warehouse and performs other tasks on the way back—a building block for flexible, intelligent production processes.
A key advantage is that many of the components required for the realization of the robot system are also used in the regional supplier industry. The proximity to companies around the automotive location of Ingolstadt could accelerate development and practical transfer.
The robot is being trained in the Laboratory for Production Engineering with the support of Jürgen Bock, Professor of AI Applications in Innovative Production and Logistics Systems, in which, virtually on powerful simulation and AI computers, it practices movement sequences, gripping techniques and navigation. This virtual training, for example, saves resources and lowers the hurdles for use in medium-sized companies.
Innovative Research by Junior Scientists
Junior scientists make an important contribution: Pauline Steffel, in particular, who works as a research associate at THI and is enrolled as a PhD candidate at THI's own Doctoral Center AI/Computer Science, deals with central questions of robot-based production processes. Among other things, together with THI students, she is researching how robots can acquire and store computer-friendly knowledge about production facilities and processes and use it to autonomously plan helpful tasks by applying sim2real reinforcement learning.
In the future, AImotion Bavaria plans to extend its setting to multi-robot scenarios and wants to investigate feasible paths to making R10-D10 more capable of learning from interaction (e.g., by using LLMs, NLP-communication, robotic foundation models). Feel free to reach out to researchers at AIMotion if you are interested in collaborating.
Working at a University of Applied Sciences
- What is a University of Applied Sciences?
- Getting to know Universities of Applied Sciences as a visiting researcher.
- Doing your PhD or becoming a professor at a University of Applied Sciences
