Robotic Intervention Learning at the Edge
A key application of edge and fog computing is that of autonomous driving. However, in order to robustly and safely maneuver around people, current approaches will need to learn at a huge scale. Scalable robot learning from seamless human-robot interaction is critical if robots are to solve a multitude of tasks in the real world. Current approaches to imitation learning suffer from one of two drawbacks. On the one hand, they rely solely on off-policy human demonstration, which in some cases leads to a mismatch in train-test distribution. On the other, they burden the human to label every state the learner visits, rendering it impractical in many applications. We argue that learning interactively from expert interventions enjoys the best of both worlds. Our key insight is that any amount of expert feedback, whether by intervention or non-intervention, provides information about the quality of the current state, the optimality of the action, or both. We formalize this as a constraint on the learner’s value function. We show how we can learn such value functions efficiently using no regret, online learning techniques. We call our approach Expert Intervention Learning (EIL), and evaluate it on a real and simulated driving task with a human expert, where it learns collision avoidance with just a few hundred samples (about one minute) of expert control.