Predictive World Models for Social Navigation


As robots begin to coexist with humans, the need for efficient and safe social robot navigation becomes increasingly pressing. In this paper we investigate how world models can enhance the effectiveness of reinforcement learning in social navigation tasks. We introduce three approaches that leverage predictive world models, which are then benchmarked against state-of-the-art algorithms. For a comprehensive and reliable evaluation, we employed multiple metrics during the training and testing phases. The key novelty of our approach consists in the integration and evaluation of predictive world models within the context of social navigation, as well as in the models themselves. Based on a diverse set of performance metrics, the experimental results provide evidence that predictive world models help improve reinforcement learning techniques for social navigation.

Publication DOI:
Divisions: College of Engineering & Physical Sciences > School of Computer Science and Digital Technologies > Applied AI & Robotics
College of Engineering & Physical Sciences > School of Computer Science and Digital Technologies
College of Engineering & Physical Sciences > Engineering for Health
College of Engineering & Physical Sciences > Aston Centre for Artifical Intelligence Research and Application
Additional Information: This version of the paper has been accepted for publication, after peer review and is subject to Springer Nature’s AM terms of use [], but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at:
Event Title: The 22nd UK Workship on Computational Intelligence
Event Type: Other
Event Location: Aston University
Event Dates: 2023-09-06 - 2023-09-08
ISBN: 978-3-031-47507-8, 978-3-031-47508-5
Last Modified: 26 Mar 2024 15:21
Date Deposited: 26 Oct 2023 11:19
Full Text Link:
Related URLs: https://link.sp ... conference/ukci (Publisher URL)
https://link.sp ... eld%20annually. (Publisher URL)
PURE Output Type: Conference contribution
Published Date: 2024-02-01
Accepted Date: 2023-07-05
Authors: Oguzie, Goodluck
Ekárt, Anikó (ORCID Profile 0000-0001-6967-5397)
Manso, Luis J. (ORCID Profile 0000-0003-2616-1120)



Version: Accepted Version

Access Restriction: Restricted to Repository staff only until 1 February 2025.

License: ["licenses_description_other" not defined]

Export / Share Citation


Additional statistics for this record