Making Uncoordinated Autonomous Machines Cooperate: A Game Theory Approach

The construction industry in Japan is facing an increasing shortage of labor that has resulted viable demands for automation and robotics. This research study applies an approach based on game theory to model cooperation in an uncoordinated autonomous environment. A set of construction machines have been designed to complete a joint task that needs cooperation without a priori knowledge of how willing or likely other agents are to cooperate. Our simulation results show that the crisped strategy model best-promoted cooperation and successfully executed the model construction task by using the robot operating system 2.