RoboFlag is an ever growing, highly dynamic testbed created at Cornell University to study and develop new algorithms for hierarchical control of teams of semi-autonomous heterogeneous vehicles. Based on the game of "Capture the Flag" the testbed offers a highly flexible environment where numerous widely applicable control problem scenarios can be recreated and studied while varying the degree of robotic autonomy and human control. The capability and diversity of the system was most recently demonstrated at an Interactive Session of the 2003 American Controls Conference where we won best presentation in our category.

The flexibility in the system has already lead the Cornell team to explore a wide range of research areas such as task allocation, artificial intelligence, primitive path planning, linear and non-linear optimization, genetic algorithm strategies, grid-based path planning, swarming artificial intelligence, streaming obstacle avoidance, geometry based sensor interpretation techniques, adaptive communication systems, and cognitive engineering, and the list continues to grow. To find out more about the testbed and the research being conducted at Cornell, please continue on to About RoboFlag.