Towards Probabilistic Planning of Explanations for Robot Navigation

TL;DR

A probabilistic framework for automated planning of explanations for robot navigation is proposed, where the preferences of different users regarding explanations are probabilistically modeled to tailor the stochasticity of the real-world human-robot interaction and the communication of decisions of the robot and its actions towards humans.

Abstract

In robotics, ensuring that autonomous systems are comprehensible and accountable to users is essential for effective human-robot interaction. This paper introduces a novel approach that integrates user-centered design principles directly into the core of robot path planning processes. We propose a probabilistic framework for automated planning of explanations for robot navigation, where the preferences of different users regarding explanations are probabilistically modeled to tailor the stochasticity of the real-world human-robot interaction and the communication of decisions of the robot and its actions towards humans. This approach aims to enhance the transparency of robot path planning and adapt to diverse user explanation needs by anticipating the types of explanations that will satisfy individual users.

Figures (3)

• Figure 1: Human preferences are represented as probabilities. Probabilities of variables that are values of the random preference variables, e.g., visual and textual for explanation representation, are instantiated such that the sum of their probabilities is 1.0 to have a valid probability distribution.
• Figure 2: Propagation of preference variables through Bernoulli functions. Robots can probabilistically track changing human preferences but also sometimes make mistakes, which more closely mimics real behavior compared to using only deterministic planning.
• Figure 3: Robot librarian starts from start_location and heads towards book_location to pick up the book. After picking up the book, it heads towards the visitor's location to hand it to the visitor. Assuming that the robot is too late, it explains its actions to the visitor while choosing visual representation, the poor level of detail, the long duration, and the global scope of its explanation while trying to respect visitor explanation preferences.