AgentChat: Multi-Agent Collaborative Logistics for Carbon Reduction

Heavy Good Vehicles (HGVs) are the second largest source of greenhouse gas emissions in transportation, after cars and taxis. However, HGVs are inefficiently utilised, with more than one-third of their weight capacity not being used during travel. We, thus, in this paper address collaborative logistics, an effective pathway to enhance HGVs' utilisation and reduce carbon emissions. We investigate a multi-agent system approach to facilitate collaborative logistics, particularly carrier collaboration. We propose a simple yet effective multi-agent collaborative logistics (MACL) framework, representing key stakeholders as intelligent agents. Furthermore, we utilise the MACL framework in conjunction with a proposed system architecture to create an integrated collaborative logistics testbed. This testbed, consisting of a physical system and its digital replica, is a tailored cyber-physical system or digital twin for collaborative logistics. Through a demonstration, we show the utility of the testbed for studying collaborative logistics.