Benchmarking the utility of Maps of Dynamics for Human-aware Motion Planning

Chittaranjan Swaminathan, Tomasz Piotr Kucner, Anna Mannucci, Martin Magnusson, Luigi Palmieri, Sergi Molina, Federico Pecora, and Achim J. Lilienthal
Frontiers in Robotics and AI

Abstract

Robots operating with humans in highly dynamic environments need not only \textit{react} to moving persons and objects but also to \textit{anticipate and adhere to} patterns of motion of dynamic agents in their environment.
Currently, robotic systems use information about dynamics locally, through tracking and predicting motion within the robot’s direct perceptual range. But doing so limits the robot’s predictive capabilities and forces it to only react to observed motion. In this paper, we explore how \emph{maps of dynamics} (MoDs) that provide information about motion patterns outside of the direct perpetual range of the robot can improve the behaviour of a robot in a dynamic environment. We formulate cost functions for four MoD representations to be used in any optimizing motion planning framework. Further, to evaluate the performance gain through using MoDs in motion planning, we have designed objective metrics and introduce a simulation framework for rapid benchmarking. We find that planners that utilize MoDs waste less time waiting for pedestrians, compared to planners that use geometric information alone. In particular, planners utilizing both intensity (proportion of observations at a grid cell w.r.t. the maximum) and direction information have better task execution efficiency

@article{swaminathan-2022-benchmarking,
title={Benchmarking the utility of Maps of Dynamics for Human-aware Motion Planning},
author={Chittaranjan Swaminathan and Tomasz Piotr Kucner and Anna Mannucci and Martin Magnusson and Luigi Palmieri and Sergi Molina and Federico Pecora and Achim J. Lilienthal},
doi={10.3389/frobt.2022.916153},
keywords={wp3},
journal={Frontiers in Robotics and AI},      
volume={9},
year=2022,
note=toappear,
}