UAS & Robotics
Google opts out of Robotics Challenge funding; DARPA expands the field
- By Kevin McCaney
- Mar 26, 2014
Team SCHAFT's robot, S-One, clears debris during the trials.
The Defense Advanced Research Projects Agency has expanded the field for the finals of its DARPA Robotics Challenge (DRC), thanks to Google’s focus on developing robotics on its own — and the company’s deep pockets.
Team SCHAFT, which came out on top of the DRC trials in December — and whose company was acquired by Google shortly before the competition — has moved to program’s self-funded track. That allows DARPA to fund two other teams for the finals, and admit a third, self-funded team that earned as many points as the other two.
“The decision by Team SCHAFT to self-fund allows DARPA to expand the competition and further develop disaster response robots,” Gill Pratt, DRC program manager, said in a statement. “This expansion is similar to what happened after DARPA held the Virtual Robotics Challenge in June 2013, when some teams shifted resources and allowed us to increase participation.”
The new finalists are teams THOR, ViGIR and KAIST, each of which scored eight out of a possible 32 points in December’s trials. (SCHAFT scored 27.).THOR and ViGIR can now get up to $500,000 in funding from DARPA; Team KAIST is self-funded.
Actually, the finals, which will be held sometime between December 2014 and June 2015 and offers a $2 million prize, will be getting a fourth extra team. The leader of Virginia Tech’s Team THOR has moved to UCLA, so its researchers have decided to split into separate teams from each school and share the funding, DARPA said.
The Tokyo-based SCHAFT Inc. is one of nine robotics companies Google has bought as part of a reported 10-year-plan to pursue robotics development, focusing on logistics and manufacturing.
But new developments in one area of robotics can help other areas. The focus of the DRC is emergency response in situations that could be too dangerous for humans. In the December trials, the robots had to drive a utility vehicle, walk through rough terrain, climb a ladder, clear debris, open a series of doors, cut through a wall, close a leaking valve and connect a fire hose.
Eight teams originally advanced to the finals. Other finalists are teams from the Florida Institute of Human & Machine Cognition, Carnegie Mellon University, MIT, NASA’s Jet Propulsion Laboratory, TRACLabs, the Worcester Polytechnic Institute and Lockheed Martin Advanced Technologies Laboratories.
Kevin McCaney is a former editor of Defense Systems and GCN.