Researchers created a deep reinforcement learning model that lets robots adapt to visual changes, maintain localization, and navigate complex indoor spaces more safely and efficiently. (Nanowerk News) ...