The pace and nature of robotics development has been remarkable indeed. Today, an automated system for any application is rarely beyond the scope of creation. Researchers from the Queen’s University, Larus Technologies and the University of British Columbia have developed an automated pool-playing gantry robot, which they believe can match humans in pool proficiency. Currently just above amateur level, the skills of the robot can be honed further to reach the level of humans. Deep Green, the pool-playing system, is a ceiling mounted robot that operates with 3 degrees of freedom. To the foot of the post is attached a spherical robotic wrist, which is the primary striking system. For two different kinds of shots (3m/s and power breaks and 12 m/s strikes), the robot uses an electromagnetic cue actuator and pneumatic cue actuator respectively. The wrist is complete with a Local Vision System (LVS) and the actual cue stick.
Deep Green relies on its Global Vision System (GVS) assisted with direction lights to play a shot. The view of the robot and the table are mapped with the help of homography, or plane mapping. Once the robot is calibrated for strikes, it takes over. Undoubtedly a huge leap in robotics, Deep Green opens up avenues for research into game companion automation. Very soon, we could be playing alongside a robot.
Via: Computer
