Last August, the CTU-CRAS team formed by our researchers and colleagues from the Department of Cybernetics was able to win among the non-sponsored teams in the DARPA Subterranean Challenge Tunnel Circuit and took 3rd place in the overall scoring.
In the upcoming part of the Sub Terranean Challenge, the DARPA Urban Circuit, which will take place in Washington (USA) on 18-27 February, we will be tackling with multi-storey terrain that includes the challenge of moving between floors on very steep and narrow stairs, possible but very difficult flythrough of vertical shafts, or new objects to detect, including the location of dangerous CO2 leakage source.
Our team is developing a multi-robotic system for operations in a previously unknown environment, for example in a disaster area. Thanks to the Defense Advanced Research Projects Agency (DARPA) competition, systems of autonomous collaborating robots can be tested, which, without human assistance, are able to explore underground spaces in conditions faithfully mimicking real action. For better 3D environmental maps, researchers have intensified collaboration with the NORLAB robotic lab from Laval University in Quebec. Colleagues from Canada have already worked with the team before, but this time they will join us personally at the competition site.
In competition with the world's best teams including MIT, ETH, JPL, UCB, CalTech, KAIST, GeorgiaTech, OregonState, UPenn or CMU, FEE CTU scientists will also be struggling for interesting financial rewards. The winner is awarded $ 500,000, the second place is $ 250,000 and the third is $ 100,000.
Both ground and flying robots of various types and sizes will be deployed in the competition. The robots are equipped with advanced sensory technology including omnidirectional cameras, stereo cameras, or lidars. The aim is to develop a system for efficient cooperation of robots with different abilities. Robots will map unknown environments and search for objects, all in a GPS-free environment with very limited communication. The greatest possible autonomy is desirable, so all the necessary calculations are done directly by the robots.
More information on CTU-CRAS participation in the DARPA competition is available on the team website.