Prospects of Language-Controlled Robotics

Language-controlled robotics offers many as yet unexploited prospects. Oier Mees, who has been nominated for the KI-Newcomer-Preis 2023 by the Federal Ministry of Education and Research, conducts research on robots and deep learning at the University of Freiburg. He is a doctoral candidate in the Autonomous Intelligent Systems research group. One question he is looking into is: How can learning robot systems acquire everyday knowledge autonomously by following natural language instructions?

Mees is trying to enable machines to interact both with the physical world and with humans in an intelligent manner and to improve in the course of time. He describes in the following video how this works in practical terms and explains the background of language-controlled robotics on the basis of deep learning.

“The ultimate goal of our research – and that of most robotics researchers, I believe – is to generalize this. That’s the biggest obstacle, the reason why the vast majority of robots today are still sitting around in labs and not in households,” says Mees of the fundamental objectives of his research.

befc7818317d13225d2031541f675ebc