Coordinate

Here we broaden the scope of our research beyond individual agents towards the coordination of behaviour of heterogeneous teams of autonomous agents to complete tasks that depend on interaction with uncertain, dynamic processes in their environment (e.g. humans, other robotic agents or environmental features). Many application domains of interest - such as our Social Care, Service & Inspection, Logistics,and Agriculture Flagships - require the coordination of such multi-agent systems. In these domains, uncertainty is present over (at least) robot and human locations in time and space, and the input of humans (e.g. willingness for, and timing of, interactions). The main challenges in this theme are creating and exploiting large-scale aggregate, probabilistic models of human-robot and robot-robot systems for coordination.
 

Congestion-Aware Policy Synthesis for Multi-Robot Systems

Street C, Pütz S, Mühlig M, Hawes N, & Lacerda B 2022 IEEE Transactions on Robotics Multi-robot systems should be robust to delays during execution. For mobile robots, one source of delays is congestion. Congestion occurs when robots arrive in the same area simultaneously, and must manoeuvre to avoid each other. Congestion may significantly increase the duration of robot navigation actions. In this paper, we present a multi-robot planning framework that reasons over the effects of congestion on navigation duration. Central to our framework is a probabilistic reservation table which summarises robot plans, capturing the effects of congestion. We then plan for each robot in turn using the congestion information in the probabilistic reservation table. We also present an iterative procedure for predicting robot performance. We evaluate our framework with extensive experiments on synthetic data and simulated robot behavior.

Multi-Unit Auctions for Allocating Chance-Constrained Resources

Gautier A, Lacerda B, Hawes N, Wooldridge M

AAAI Conference on Artificial Intelligence (AAAI 2023)

Sharing scarce resources is a key challenge in multi-agent interaction, especially when individual agents are uncertain about their future consumption. We present a new auction mechanism for preallocating multi-unit resources among agents, while limiting the chance of resource violations. By planning for a chance constraint, we strike a balance between worst-case approaches, which under-utilise resources, and expected-case approaches, which lack formal guarantees.

We also present an algorithm that allows agents to generate bids via multi-objective reasoning, which are then submitted to the auction. We then discuss how the auction can be extended to non-cooperative scenarios. Finally, we demonstrate empirically that our auction outperforms state-of-the-art techniques for chance-constrained multi-agent resource allocation in complex settings with up to hundreds of agents.