Learning From Demonstrations Using Signal Temporal Logic in Stochastic and Continuous Domains

Learning control policies that are safe, robust and interpretable are prominent challenges in developing robotic systems. Learning-from-demonstrations with formal logic is an arising paradigm in reinforcement learning to estimate rewards and extract robot control policies that seek to overcome these challenges. In this approach, we assume that mission-level specifications for the robotic system are expressed in a suitable temporal logic such as Signal Temporal Logic (STL). The main idea is to automatically infer rewards from user demonstrations (that could be suboptimal or incomplete) by evaluating and ranking them w.r.t. the given STL specifications. In contrast to existing work that focuses on deterministic environments and discrete state spaces, in this letter, we propose significant extensions that tackle stochastic environments and continuous state spaces.

Automated summary

Learning control policies that are safe, robust and interpretable is a major challenge in developing robotic systems. Using formal logic in learning-from-demonstrations is a new paradigm in reinforcement learning to estimate rewards and extract robot control policies to overcome these challenges. The approach involves inferring rewards automatically from user demonstrations based on specified logic, ranking them to determine optimal control policies.