EUVIMEDEUVIMED entdecken
Crossref

Heterogeneous Policy Networks for Composite Robot Team Communication and Coordination

Esmaeil Seraj, Rohan Paleja, Luis Pimentel, Kin Man Lee, Zheyuan Wang, Daniel Martin, Matthew Sklar, John Zhang, Zahi Kakish, Matthew Gombolay

IEEE Transactions on Robotics · 2024 · 40 · 3833-3849

DOI: 10.1109/tro.2024.3431829 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Robotics, 2024. Verfasst von Esmaeil Seraj und 9 weiteren Autor:innen. Die Arbeit wurde bisher 8-mal zitiert (Crossref) und nennt 73 Literaturangaben. Gefördert von: Naval Research Lab, Lockheed Martin Corporation, Office of Naval Research.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

High-performing human–human teams learn intelligent and efficient communication and coordination strategies to maximize their joint utility. These teams implicitly understand the different roles of heterogeneous team members and adapt their communication protocols accordingly. Multiagent reinforcement learning (MARL) has attempted to develop computational methods for synthesizing such joint coordination–communication strategies, but emulating heterogeneous communication patterns across agents with different state, action, and observation spaces has remained a challenge. Without properly modeling agent heterogeneity, as in prior MARL work that leverages homogeneous graph networks, communication becomes less helpful and can even deteriorate the team's performance. In the past, we proposed heterogeneous policy networks (HetNet) to learn efficient and diverse communication models for coordinating cooperative heterogeneous teams. In this extended work, we extend HetNet to support scaling heterogeneous robot teams. Building on heterogeneous graph-attention networks, we show that HetNet not only facilitates learning heterogeneous collaborative policies, but also enables end-to-end training for learning highly efficient binarized messaging. Our empirical evaluation shows that HetNet sets a new state-of-the-art in learning coordination and communication strategies for heterogeneous multiagent teams by achieving an 5.84% to 707.65% performance improvement over the next-best baseline across multiple domains while simultaneously achieving a 200× reduction in the required communication bandwidth.

Förderung

  • Naval Research Lab
  • Lockheed Martin Corporation
  • Office of Naval Research
  • Sandia National Laboratories

Zitieren

APA

Seraj, E., Paleja, R., Pimentel, L., Lee, K. M., Wang, Z., Martin, D., Sklar, M., Zhang, J., Kakish, Z., & Gombolay, M. (2024). Heterogeneous Policy Networks for Composite Robot Team Communication and Coordination. IEEE Transactions on Robotics, 40, 3833-3849. https://doi.org/10.1109/tro.2024.3431829

Vancouver

Seraj E, Paleja R, Pimentel L, Lee KM, Wang Z, Martin D, et al. Heterogeneous Policy Networks for Composite Robot Team Communication and Coordination. IEEE Transactions on Robotics. 2024;40:3833-3849. doi:10.1109/tro.2024.3431829

BibTeX
@article{Seraj2024,
  author = {Seraj, Esmaeil and Paleja, Rohan and Pimentel, Luis and Lee, Kin Man and Wang, Zheyuan and Martin, Daniel and Sklar, Matthew and Zhang, John and Kakish, Zahi and Gombolay, Matthew},
  title = {Heterogeneous Policy Networks for Composite Robot Team Communication and Coordination},
  journal = {IEEE Transactions on Robotics},
  year = {2024},
  volume = {40},
  pages = {3833--3849},
  doi = {10.1109/tro.2024.3431829}
}
Kostenlos starten →