Role-based multi-robot warehouse coordination with reinforcement learning assisted task selection
4 affiliations
- 1Electronic Information & Interdisciplinary Research Institute, Yango University
- 2College of Artificial Intelligence and, Fujian University Engineering Research Center of Spacial Data Mining and Application, Yango University
- 3Yango University
- 4School of Software, Quanzhou University of Information Engineering
Abstract
We study decentralised coordination in large-scale RMFS, where capacity-limited fleets fulfil orders under stochastic demand, congestion, and real-time constraints. We formalise this problem as a role-structured decentralised partially observable Semi-Markov decision process that extends the semi-Markov multi-agent framework to include role-conditioned macro-actions with role-dependent eligibility constraints and the quorum role switching mechanism for density-driven, decentralised role transitions. Unlike generic macro-action initiation sets, our eligibility gates enforce physical feasibility by dynamically disabling roles based on instantaneous load, capacity, and shelf-assignment state, while the quorum role switching balances local task demand against agent supply without centralised coordination. To optimise within this structured action space, we introduce HSC-Net, a hybrid actor–critic architecture that combines role-specific heuristic candidate generation with convolutional spatial encoding and cross-attention-based candidate scoring, supported by a centralised graph-attention critic over the warehouse sector graph. Our framework enables asynchronous multi-hop relay without fixed infrastructure or agent rendezvous, decoupling intermediate shelf transport from direct delivery. Extensive simulations across different warehouse scales demonstrate significant improvements in throughput, completion time, and congestion balance against rule-based heuristics and value-based deep Q-learning baselines.
Related
- Continuous control deep reinforcement learning for joint resource optimization in STAR-RIS assisted mmWave-NOMA systems with hardware impairments
- Rhythm-centred music interventions for motor coordination and rehabilitation: a scoping review of mechanisms, design principles, and technology-assisted assessment.
- Reinforcement learning–coupled neural ode modelling reveals ferroptotic resolution failure as a shared dynamical axis linking periodontitis and coronary atherogenesis
- Reinforcement learning as a foundation for psychodynamics
- Multiproperty optimization of therapeutic antibodies through sequence editing via large language model and reinforcement learning
BibTeX
@article{nantogma2026rolebased,
title = {{Role-based multi-robot warehouse coordination with reinforcement learning assisted task selection}},
author = {Sulemana Nantogma and Yakubu Imrana and Xiaorong Xu and Isaac Amankona Obiri and Anhui Liang},
journal = {Scientific Reports},
year = {2026},
month = aug,
doi = {10.1038/s41598-026-63868-3},
url = {https://doi.org/10.1038/s41598-026-63868-3},
note = {CC BY},
}
RIS
TY - JOUR
TI - Role-based multi-robot warehouse coordination with reinforcement learning assisted task selection
AU - Sulemana Nantogma
AU - Yakubu Imrana
AU - Xiaorong Xu
AU - Isaac Amankona Obiri
AU - Anhui Liang
PY - 2026
DA - 2026/08/01
JO - Scientific Reports
DO - 10.1038/s41598-026-63868-3
UR - https://doi.org/10.1038/s41598-026-63868-3
AB - We study decentralised coordination in large-scale RMFS, where capacity-limited fleets fulfil orders under stochastic demand, congestion, and real-time constraints. We formalise this problem as a role-structured decentralised partially observable Semi-Markov decision process that extends the semi-Markov multi-agent framework to include role-conditioned macro-actions with role-dependent eligibility constraints and the quorum role switching mechanism for density-driven, decentralised role transitions. Unlike generic macro-action initiation sets, our eligibility gates enforce physical feasibility by dynamically disabling roles based on instantaneous load, capacity, and shelf-assignment state, while the quorum role switching balances local task demand against agent supply without centralised coordination. To optimise within this structured action space, we introduce HSC-Net, a hybrid actor–critic architecture that combines role-specific heuristic candidate generation with convolutional spatial encoding and cross-attention-based candidate scoring, supported by a centralised graph-attention critic over the warehouse sector graph. Our framework enables asynchronous multi-hop relay without fixed infrastructure or agent rendezvous, decoupling intermediate shelf transport from direct delivery. Extensive simulations across different warehouse scales demonstrate significant improvements in throughput, completion time, and congestion balance against rule-based heuristics and value-based deep Q-learning baselines.
SN - 2045-2322
ER -