The Ulm Sparrows: Research into Sensorimotor Integration, Agency Learning, and Multiagent Cooperation

We describe the motivations, research issues, current results, and future directions of The Ulm Sparrows, a project that aims at the design and implementation of a team of robotic soccer players. The RoboCup challenge was proposed by Kitano et al..1] in order to revive research in AI and to provide the eld with a new long-range research challenge. Creating a powerful team of skilled robotic soccer players involves much more than enhancing and integrating AI technology. Mechatronic devices, specialized hardware for controlling sensors and actuators, control theory, sensor interpretation and fusion, neural networks, evolutionary computation, vision, multiagent cooperation are examples for elds strongly involved with the RoboCup challenge. Thus, engineering a team of robotic soccer players that is able to successfully compete in championship games requires truely interdisciplinary research and the eeective cooperation of many disciplines. This is our main scientiic reason for setting up a local research eeort related to RoboCup and to implement The Ulm Sparrows, a team of soccer agents that is currently projected to participate both in the mid-size real robot league as well as the simulation league. A secondary reason is the coincidence of RoboCup research issues with mid-term research goals of SMARTT2] 1 , a local, large, interdisciplinary joint research project that aims at the development of a complete cognitive architecture for an adaptive robotic agent performing complex tasks. Not the least reasons for starting RoboCup activities were their attractiveness to students and its potential for appealing educational activities. 1 See www.uni-ulm.de/SMART/ for more information.