Using policy gradient reinforcement learning on autonomous robot controllers | IEEE Conference Publication | IEEE Xplore