Analyzing Human Models that Adapt Online

TL;DR

This paper introduces a dynamical systems approach to analyze online-adapting human models in robotics, enabling safety and learning guarantees through reachability and optimal control tools.

Contribution

It presents a novel framework modeling the robot's learning process as a dynamical system, allowing analysis of learning speed and safety guarantees.

Findings

Can compute the set of hypotheses the robot can learn in finite time.

Provides worst-case and best-case learning time estimates.

Validated in four human-robot interaction domains.

Abstract

Predictive human models often need to adapt their parameters online from human data. This raises previously ignored safety-related questions for robots relying on these models such as what the model could learn online and how quickly could it learn it. For instance, when will the robot have a confident estimate in a nearby human's goal? Or, what parameter initializations guarantee that the robot can learn the human's preferences in a finite number of observations? To answer such analysis questions, our key idea is to model the robot's learning algorithm as a dynamical system where the state is the current model parameter estimate and the control is the human data the robot observes. This enables us to leverage tools from reachability analysis and optimal control to compute the set of hypotheses the robot could learn in finite time, as well as the worst and best-case time it takes to learn them. We demonstrate the utility of our analysis tool in four human-robot domains, including autonomous driving and indoor navigation.