A Decision-Making Algorithm for an Air-Hockey Robot that Decides Actions Depending on its Opponent Player's Motions

In this study, we propose a novel air-hockey robot that can select the optimal actions according to the behavior of an opponent human player. The attack behavior of the robot is optimized during an air hockey game by recognizing the motions of the puck and the human hand. First, we explain how to calculate the attack position. Second, we explain the algorithm for decision making using an attack value based on the puck and the position of the human hand. The attack value is calculated by considering the physical constraints of both the attack and defense sides. Then, we show experimental data and explain the reason for robot behaviors based on the calculated attack value.