The Dutch are (Robot) Football Champions of the World!
The World Cup Football in Russia has entered the final stages of the tournament. Meanwhile the Netherlands are already football world champions! Dutch Technical University Eindhoven (TU/e) won the Robot Football World Cup (RoboCup) for the fourth time.
Image: ©Holland Branding / Holland Toolkit
Tech United, as the TU/e team is called, won the RoboCup final in Montreal by defeating the Portuguese side, Cambada, 1-0. The robots were designed and built by the university, as was the software.
The Tech United team consists of 5 fully automonous robots and a couple of dozen students. The group competes against other university teams in two 15 minutes halves on a battle field measuring 18 by 12 meters.
RoboCup’s ultimate goal is to beat a human football team by 2050.
Tech United is also involved in the development of service robots. Their service robots AMIGO and SERGIO participated in the @Home competition. During the event the robots have to complete tasks like taking something out of the kitchen and recognize people. In the near future these robots will be used to help the elderly to function independently .
Service robots and football robots share the same Robo-DNA. Many parts in the football robot are also used in the service robot, like the wheels and knowledge transfer modules. They also use the same sophisticated software, programed to communicate in a continuously changing environment.
Development acceleration by sharing information
The goal of the competitions is to accelerate the development of robots and Artificial Intelligence (AI). The universities are required to share their technical specs and source codes with the other teams after the contest.
The robots are part of the RoboEarth project, their goal is to let the robots learn from each other. Object recognition, navigation, and other activities are exchanged via the RoboEarth network.
The AMIGO robot for instance, makes contact with the RoboEarth database autonomously and downloads the necessary instructions that are needed to serve a particular product - such as fruit juice.
Through the network, RoboEarth robots will be able to deal with unforeseen circumstances, instead of programming every conceivable situation. Since the robots learn from each event and share their experiences, they will respond faster and improve their ability to learn and to react in new situations.