One key assumption of Belief Space Planning (BSP) is that the data association is known perfectly. In this paper, we relax this assumption in the context of non-myopic planning as well as belief being a Gaussian Mixture Model (GMM). Interestingly, explicit reasoning about the data association within the belief enables our framework to have parsimonious data association, thereby resulting in a scalable solution compared with naïve permutational approaches. Unlike in some of the recent approaches where the number of components in a GMM belief can only be reduced, in our approach this can also go up such as due to perceptual aliasing present in the environment. Furthermore, our approach naturally integrates with inference, providing a unified framework for robust passive and active perception. We demonstrate key aspects of our approach and its comparison with the state of the art on a general abstract domain as well as in a real robot setup.

This paper argues in favour of using formal methods to ensure safety of deployed stochastic policies learned by robots in unstructured environments. It has been demonstrated that multi-objective learning alone is not sufficient to ensure globally safe behaviours in such robots, whereas learning-specific methods yield deterministic policies which are less flexible or effective in practice. Under certain restrictions on state-space, modelling safety using probabilistic computational tree logic and ensuring such safety via automated repair can overcome these shortcomings. Promising results are obtained on a realistic setup and pros and cons of such method are discussed.

Given a stochastic policy learned by reinforcement, we wish to ensure that it can be deployed on a robot with demonstrably low probability of unsafe behavior. Our case study is about learning to reach target objects positioned close to obstacles, and ensuring a reasonably low collision probability. Learning is carried out in a simulator to avoid physical damage in the trial-and-error phase. Once a policy is learned, we analyze it with probabilistic model checking tools to identify and correct potential unsafe behaviors. The whole process is automated and, in principle, it can be integrated step-by-step with routine task-learning. As our results demonstrate, automated fixing of policies is both feasible and highly effective in bounding the probability of unsafe behaviors.

In this paper we consider the problem of ensuring that a multi-agent robot control system is both safe and effective in the presence of learning components. Safety, i.e., proving that a potentially dangerous configuration is never reached in the control system, usually competes with effectiveness, i.e., ensuring that tasks are performed at an acceptable level of quality. In particular, we focus on a robot playing the air hockey game against a human opponent, where the robot has to learn how to minimize opponent's goals (defense play). This setup is paradigmatic since the robot must see, decide and move fastly, but, at the same time, it must learn and guarantee that the control system is safe throughout the process. We attack this problem using automata-theoretic formalisms and associated verification tools, showing experimentally that our approach can yield safety without heavily compromising effectiveness.