Improving Local Motion Planning with a Reinforcement Learning Approach | IEEE Conference Publication | IEEE Xplore