IfM's Supply Chain AI Lab at AI UK 2023
The Supply Chain AI Lab (SCAIL) at the IfM recently participated at AI UK 2023 in London – the UK’s national showcase of data science and artificial intelligence – with a demo of an innovative supply system which aims to significantly lower emissions from delivery vehicles.
Dr Liming Xu explains: “The demo, ‘AgentChat: Agent-based Collaborative Logistics for Carbon Reduction’, presented as part of the environment sector of AI UK 2023 showcases a collaborative logistics platform using self-driving cars.
"Here, intelligent software agents that act on behalf of carriers, customers, and suppliers discover and negotiate truck-sharing opportunities using AI algorithms, share monetary value and converge on optimal delivery routes. A physical map and its digital dashboard were developed and connected as a digital twin to visualise this cyber-physical system.”
In the real world, this system can be applied to help delivery vehicles operate more sustainably and efficiently. British roads are teeming with vehicles that transport goods between providers and consumers, and heavy goods vehicles (HGVs) are currently responsible for 19% of domestic transport emissions in the UK. Because of a lack of high-level coordination, many HGV journeys take place without much cargo, with average truck utilisation remaining around 60%.
It is both efficient and sustainable to optimise the use of available space in delivery vehicles, but it is not widely done due to lack of orchestration, and the transaction cost involved in available solutions, which create a lock-in effect on carriers. Comprehensive interventions have the potential to be very impactful, given the size of the delivery industry which connects the dots of many of our supply chains. A system like the one the SCAIL team is working on can contribute to greater collaboration between delivery services – a bit like carpooling.
Professor Alexandra Brintrup said “We urgently need a radical re-think of logistics systems in order to be able to achieve net-zero. Current delivery planning systems cannot optimise routes at the system level due to a lack of information sharing. Automation can help alleviate the manual effort involved in truck sharing, while at the same time allowing all parties to gain value from it. But logistics providers will need to get used to a new way of thinking. ”
The demo will continue to be housed at the IfM, and the research group plans to develop its capabilities further with more advanced features.