We focus on learning robot objective functions from human guidance: specifically, from physical corrections provided by the person while the robot is acting. Objective functions are typically parametrized in terms of features, which capture aspects of the task that might be important. When the person intervenes to correct the robot's behavior, the robot should update its understanding of which features matter, how much, and in what way. Unfortunately, real users do not provide optimal corrections that isolate exactly what the robot was doing wrong. Thus, when receiving a correction, it is difficult for the robot to determine which features the person meant to correct, and which features were changed unintentionally. In this paper, we propose to improve the efficiency of robot learning during physical interactions by reducing unintended learning. Our approach allows the human-robot team to focus on learning one feature at a time, unlike state-of-the-art techniques that update all features at once. We derive an online method for identifying the single feature which the human is trying to change during physical interaction, and experimentally compare this one-at-a-time approach to the all-at-once baseline in a user study. Our results suggest that users teaching one-at-a-time perform better, especially in tasks that require changing multiple features.

One of the most difficult challenges in robot motion planning is to account for the behavior of other moving agents, such as humans. Commonly, practitioners employ predictive models to reason about where other agents are going to move. Though there has been much recent work in building predictive models, no model is ever perfect: an agent can always move unexpectedly, in a way that is not predicted or not assigned sufficient probability. In such cases, the robot may plan trajectories that appear safe but, in fact, lead to collision. Rather than trust a model’s predictions blindly, we propose that the robot should use the model’s current predictive accuracy to inform the degree of confidence in its future predictions. This model confidence inference allows us to generate probabilistic motion predictions that exploit modeled structure when the structure successfully explains human motion, and degrade gracefully whenever the human moves unexpectedly. We accomplish this by maintaining a Bayesian belief over a single parameter that governs the variance of our human motion model. We couple this prediction algorithm with a recently proposed robust motion planner and controller to guide the construction of robot trajectories that are, to a good approximation, collision-free with a high, user-specified probability. We provide extensive analysis of the combined approach and its overall safety properties by establishing a connection to reachability analysis, and conclude with a hardware demonstration in which a small quadcopter operates safely in the same space as a human pedestrian.

Abstract-In order to safely operate around humans, robots can employ predictive models of human motion. Unfortunately, these models cannot capture the full complexity of human behavior and necessarily introduce simplifying assumptions. As a result, predictions may degrade whenever the observed human behavior departs from the assumed structure, which can have negative implications for safety. In this paper, we observe that how "rational" human actions appear under a particular model can be viewed as an indicator of that model's ability to describe the human's current motion. By reasoning about this model confidence in a real-time Bayesian framework, we show that the robot can very quickly modulate its predictions to become more uncertain when the model performs poorly. Building on recent work in provably-safe trajectory planning, we leverage these confidence-aware human motion predictions to generate assured autonomous robot motion. Our new analysis combines worst-case tracking error guarantees for the physical robot with probabilistic time-varying human predictions, yielding a quantitative, probabilistic safety certificate. We demonstrate our approach with a quadcopter navigating around a human.

Real-world autonomous vehicles often operate in a priori unknown environments. Since most of these systems are safety-critical, it is important to ensure they operate safely in the face of environment uncertainty, such as unseen obstacles. Current safety analysis tools enable autonomous systems to reason about safety given full information about the state of the environment a priori. However, these tools do not scale well to scenarios where the environment is being sensed in real time, such as during navigation tasks. In this work, we propose a novel, real-time safety analysis method based on Hamilton-Jacobi reachability that provides strong safety guarantees despite environment uncertainty. Our safety method is planner-agnostic and provides guarantees for a variety of mapping sensors. We demonstrate our approach in simulation and in hardware to provide safety guarantees around a stateof-the-art vision-based, learning-based planner. Videos of our approach and experiments are available on the project website 1 .

No abstract