A leader-follower formation control using deep reinforcement learning environment, In which every agent can learn to follow the leader agent by keeping track of a certain distance to that leader, avoiding obstacles, and avoiding collision with the other agents.