CaltechAUTHORS
  A Caltech Library Service

Adaptive Conformal Prediction for Motion Planning among Dynamic Agents

Dixit, Anushri and Lindemann, Lars and Wei, Skylar X. and Cleaveland, Matthew and Pappas, George J. and Burdick, Joel W. (2022) Adaptive Conformal Prediction for Motion Planning among Dynamic Agents. . (Unpublished) https://resolver.caltech.edu/CaltechAUTHORS:20221219-234025206

[img] PDF - Submitted Version
Creative Commons Attribution.

464kB

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20221219-234025206

Abstract

This paper proposes an algorithm for motion planning among dynamic agents using adaptive conformal prediction. We consider a deterministic control system and use trajectory predictors to predict the dynamic agents' future motion, which is assumed to follow an unknown distribution. We then leverage ideas from adaptive conformal prediction to dynamically quantify prediction uncertainty from an online data stream. Particularly, we provide an online algorithm uses delayed agent observations to obtain uncertainty sets for multistep-ahead predictions with probabilistic coverage. These uncertainty sets are used within a model predictive controller to safely navigate among dynamic agents. While most existing data-driven prediction approached quantify prediction uncertainty heuristically, we quantify the true prediction uncertainty in a distribution-free, adaptive manner that even allows to capture changes in prediction quality and the agents' motion. We empirically evaluate of our algorithm on a simulation case studies where a drone avoids a flying frisbee.


Item Type:Report or Paper (Discussion Paper)
Related URLs:
URLURL TypeDescription
http://arxiv.org/abs/2212.00278arXivDiscussion Paper
ORCID:
AuthorORCID
Lindemann, Lars0000-0003-3430-6625
Wei, Skylar X.0000-0002-6336-9433
Cleaveland, Matthew0000-0003-3671-5297
Pappas, George J.0000-0001-9081-0637
Burdick, Joel W.0000-0002-3091-540X
Additional Information:© 2023 A. Dixit, L. Lindemann, S.X. Wei, M. Cleaveland, G.J. Pappas & J.W. Burdick. Attribution 4.0 International (CC BY 4.0). Lars Lindemann, Matthew Cleaveland, and George J. Pappas were generously supported by NSF award CPS-2038873. The work of Anushri Dixit and Skylar Wei was supported in part by DARPA, through the LINC program.
Funders:
Funding AgencyGrant Number
NSFECCS-2038873
Defense Advanced Research Projects Agency (DARPA)UNSPECIFIED
Record Number:CaltechAUTHORS:20221219-234025206
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20221219-234025206
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:118457
Collection:CaltechAUTHORS
Deposited By: George Porter
Deposited On:20 Dec 2022 03:39
Last Modified:20 Dec 2022 03:39

Repository Staff Only: item control page