The DARPA Robotics Challenge

During this year, a competition between robots from all over the world was held. It aimed to find the best one in case of emergency

Some years ago, the DARPA agency had proposed a challenge to the technological centers from all over the world: they had to create a robot which could operate without human help whenever it would be necessary, like in natural or nuclear disasters. 25 teams used their best technologies, together with hard researches, to build that machine.

theroboticschallenge.org

The challenge was held on 5th and 6th June 2015, in which the robots had to perform 8 tasks. They had to walk and interract with objects, but they also had to drive a car, close a valve and other things. The ranking was determined by the number of tasks accomplished and the time spent, with winning prizes of 2 million dollars, 1 million dollars and 500 thousands dollars respectively for 1st, 2nd and 3rd places. The winner has been the team KAIST from South Corea with its robot DRC-Hubo (in the photo), which succeed in all the tasks with the time of 44:28. At the second and third place, two US teams, the team IHMC Robotics and the Tartan Rescue team. Only these three teams’ robots were able to complete all of the paths, but the result was very good: the development of these machines allows us to think that we can be protected or saved by a robot in the near future.

Italy in the challenge

theroboticschallenge.org

Two of the best italian research centers, Istituto Italiano di Tecnologia and the university center E. Piaggio in Pisa, participated at DARPA Robotics Challenge building Walk-Man. This humanoid robot, 185 cm high and 120 kg weight, ranked 17th completing only 2 task in a time of 36:35, because of a battery problem. But it isn’t so bad, because these italian centers have demonstrated their ability building Walk-Man in only ten months.

27 people, led by a robotic engineer, worked on the robot. Walk-Man can move and operate as an human, because of its intelligence, due to two computers with INTEL i7 quadcore processors which control the robot view (thanks to cameras and a tridimensional laser), perception (using control boards, sensors and accelerometers) and motion (with 33 engine combining the various joints).