Conceived in 2008, the AcYut project had a goal of developing a humanoid for industrial and disastrous environments. Six humanoids have been created by the project and each new robot has increased in complexity. The first robot, AcYut 1, was a remote controlled humanoid that won 6th place in the 2008 world RoboGames. The current AcYut 6 is an autonomous intelligent humanoid with coupling constructs that allow it to be driven by multiple motors.
Image processing uses the fish-eye lens due to their wide field of vision. To obtain the robot’s global position, the system uses Monte Carlo localization. Footstep planning maps the given path into footstep positions. On a circular path the robot slows down and on a straight path it speeds up to maximum speed. In case the camera does not offer enough landmarks to assure a confident global localization of the robot, the system has a motion model that provides the best estimate of the position based on the history of the robot’s moves.
The AcYut is looking into reducing the weight of their robots by using Carbon Fibre Reinforced Plastic as the building material as well as including springs to facilitate jumping and running. In addition, the team is planning on integrating a cooperative behavioral framework to allow collaboration between robots.