University of Bonn Household Robots Win the German Open

The University of Bonn’s NimbRo team - earned the top score in the @Home l
The University of Bonn’s NimbRo team - earned the top score in the @Home league for household robots at the RoboCup German Open in Kassel. Photo: Autonomous Intelligent Systems/University of Bonn all’images in original size .
The University of Bonn’s Team NimbRo was the top scoring team in the @Home league for household robots at the RoboCup German Open, held in Kassel from April 18-20. Developed by the Autonomous Intelligent Systems working group at the Institute of Computer Science, NimbRo’s lifelike soccer-playing robots had the best final-round test results, according to the judges panel. The robots are able to navigate around on their own, grasp and place objects and interact with people via voice dialogue technology.

The RoboCup German Open is Europe’s largest annual competition for intelligent autonomous service robots in the @Home league. The purpose of the competition is to promote the development of assistive robots for everyday living environments, to help assisted-living patients, for example. For the competition, a series of tests are conducted in a realistic home environment that allow evaluation of the robots’ capabilities. The capabilities principally evaluated are intuitive human-robot interaction through speech and body language, rapid and safe navigation in unstructured everyday environments and object handling.

The many different tasks to be completed in the robot trials include greeting guests, enforcing house rules, putting away groceries, setting the breakfast table, serving guests in a restaurant and cleaning up the kitchen. The team from Bonn programmed a mobile robot with a flexible chassis and two arms to perform these tasks, which perceives its surroundings via cameras, laser scanners and a microphone. The robot is equipped with 23 separate motors and a speaker to enable it to independently carry out assistive tasks. Methods derived from artificial intelligence research are deployed, including image understanding, speech understanding, motion planning and dialogue systems.

In the very first round of testing the University of Bonn team took the lead, and held onto it throughout the three-day competition. For the final, the top two teams demonstrated performance of a task chosen by the respective teams itself. The University of Bonn robot showed how it navigates around an apartment, comprehends language and gestures and supports users preparing dinner by recognizing food ingredients on hand and offering relevant recipe suggestions. The judges panel, consisting of select experts and leaders of the other teams, found this a convincing performance.

"Assistive robots are a key development in enabling assisted-living patients to continue living on their own in their familiar home environment," says Professor Sven Behnke, who is head of the Autonomous Intelligent Systems working group and Director of the University of Bonn Institute of Computer Science, section VI - Intelligent Systems and Robotics. "The competition in Kassel provided important insights for our future research work," adds Professor Behnke, who is also a member of the transdisciplinary research areas Modelling and Sustainable Futures, the PhenoRob Cluster of Excellence, the Lamarr Institute for Machine Learning and Artificial Intelligence and the Center for Robotics.