We are interested in the possibility of "self-driving" supply chains, in which human and digital software agents work together make predictive and prescriptive decisions to controls the flow of materials, information and finance.
This proposition needs connected supply chain technologies across planning, procurement, manufacturing and logistics that work beyond an organisation’s four walls. Examples include autonomous response to changes in customer demand and supply availability across the entire value chain to collectively adjust resource and distribution plans and production schedules. Automation at this level can not only alleviate the burden on human supply chain professionals, but also make significant leaps in areas where manual orchestration is not possible.
Our initial work in this domain started from the need to plan for frequent, low level service events, where part replacement signals propagated across the value chain to autonomously select suppliers, procure goods and agree on pricing.
Follow on work through project PitchIn, focussed on creating demonstrative platforms to a response to a traffic disturbance or a fridge break down during the cold chain delivery, for example through automated rerouting. We created automated pricing algorithms to adjust prices based on delivery delays, and resulting quality degradations and monitoring of supplier delivery performance. A description of our demonstrator activity can be accessed here.
The concept of autonomous supply chains cuts across a number of related topics, including cooperative multi-agent systems, automation of supply chain operations, collective and federated learning systems, and tools such as IoT and Supply Chain Digital Twins.
Our current projects in this Theme include:
ChattyTwins, sponsored by Alan Turing Institute, where we investigate how digital twins across members of supply chain can communicate to optimise operations,
Agent Chat, sponsored by EPSRC, where we are building autonomous freight carrier agents to identify opportunities to share heavy good vehicles, and negotiate on how the resulting value can be shared
Funded by: Alan Turing Institute, EPSRC, Research England
Researchers: Liming Xu, Stefan Schoepf, Stephen Mak
Partners: Fetch.ai, VCL, Accenture