Information Gathering Actions Over Human Internal State

Dorsa Sadigh, Shankar Sastry, Sanjit A. Seshia, and Anca D. Dragan. Information Gathering Actions Over Human Internal State. In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 66–73, October 2016.
Best Paper Finalist (cognitive robotics category). Featured in IEEE Spectrum.

Download

[pdf] 

Abstract

Much of estimation of human internal state (goal, intentions, activities, preferences, etc.) is <i>passive</i>: an algorithm observes human actions and updates its estimate of human state. In this work, we embrace the fact that robot actions affect what humans do, and leverage it to improve state estimation. We enable robots to do <i>active information gathering</i>, by planning actions that probe the user in order to clarify their internal state. For instance, an autonomous car will plan to nudge into a human driver's lane to test their driving style. Results in simulation and in a user study suggest that active information gathering significantly outperforms passive state estimation.

BibTeX

@inproceedings{sadigh-iros16,
  author    = {Dorsa Sadigh and Shankar Sastry and Sanjit A. Seshia and Anca D. Dragan},
  title     = {Information Gathering Actions Over Human Internal State},
  booktitle = {Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  month = "October",
  year = {2016},
  pages = {66--73},
  abstract = {Much of estimation of human internal state (goal, intentions, activities, preferences, etc.) is \emph{passive}: an algorithm observes human actions and updates its estimate of human state. In this work, we embrace the fact that robot actions affect what humans do, and leverage it to improve state estimation. We enable robots to do \emph{active information gathering}, by planning actions that probe the user in order to clarify their internal state. For instance, an autonomous car will plan to nudge into a human driver's lane to test their driving style. Results in simulation and in a user study suggest that active information gathering significantly outperforms passive state estimation.},
  wwwnote = {<b>Best Paper Finalist</b> (cognitive robotics category). Featured in <a href="http://spectrum.ieee.org/cars-that-think/transportation/human-factors/autonomous-cars-could-determine-your-driving-style-by-gently-probing-you">IEEE Spectrum</a>.}
}