DRC-Hubo humanoid robot wins DARPA Robotics Challenge



    The South Korean KAIST team, which created the DRC-Hubo robot , will take away the main prize of $ 2 million as the winner of the DARPA Robotics Challenge , which was held on Friday and Saturday in California.

    This robot coped better than anyone else with the passage of an obstacle course, which included tests for opening doors, unscrewing valves, climbing stairs, driving a car, drilling holes in the wall, and so on.

    The entire obstacle course is shown schematically in the illustration.



    In total, 25 robots participated in the competition. The second and third places were taken by the IHMC Robotics and Tartan Rescue teams, both from the USA. They received prizes of $ 1 million and $ 500 thousand, respectively.

    But much more important than money is a sense of triumph: victory in the prestigious DARPA contest, which lasted more than two years. The Pentagon announced the competition in early 2013, after which two preliminary rounds of the competition took place.



    During the two days of the final competition on June 5-6, each robot was given two attempts to pass the obstacle course, which includes eight tasks. These tasks are specially selected taking into account the possible actions of the rescue robot after a man-made disaster, such as in Fukushima.

    Of the 24 participating robots, 12 were from the USA, five from Japan, three from South Korea, two from Germany, one from Italy and Hong Kong.

    But the victory went to the South Korean robot Hubo ("HUmanoid roBOt"), which was especially impressed with its ability to work on its knees. He has special wheels on his lap, so the robot can even ride in the right direction.



    All three winners successfully completed eight tasks, but Hubo did it 6 minutes faster than the IHMC Robotics and 10 minutes faster than the CHIMP from Tartan Rescue.

    Hubo has been developed since 2002, it weighs 80 kg with a height of 180 cm. Sensors and software allow it to form a three-dimensional map of the area and recognize objects. For the competitive version of DRC-Hubo, the developers completely rewrote the walking algorithm and made the robot more stable.

    Not all contestants managed to overcome the obstacle course; some failed or fell (see video).



    In the future, such robots will be able to completely replace humans during rescue operations after accidents like Fukushima.

    Also popular now: