Development of Warehouse Robot Arms for Grasping Objects

Demand for robots is generally growing at a rapid clip, especially for the warehouse automation. Our finial goal is to develop a robot that can assist in grasping objects in the warehouse. The first step is to make the robot arm reach the target position accurately, and then grasp, which is also the goal for this project. We mainly used deep deterministic policy gradient (DDPG) and some other relative algorithms to train the robot arm.