Transforming manufacturing with autonomous supply chains
Digital technologies offer huge opportunities for supply chain transformation. Professor Alexandra Brintrup from the Institute for Manufacturing (IfM), explains how agent-based systems that use AI and the Internet of Things (IoT) are helping to make supply chains smarter and more efficient.
Supply chains can be large scale and complex, often consisting of multiple interacting companies that don’t always share the same information systems. This can lead to a lack of real-time, traceable information across the supply chain, making operations inefficient and making them vulnerable to coordination challenges and disruptions (as seen during COVID-19).
Without visibility of this kind of information, it is difficult to plan for, or respond to, problems such as volatile market demand, supplier delays or driver shortages.
However, new technologies such as AI and IoT offer the opportunity for companies to transform supply chains into autonomous supply chain ecosystems, resulting in more seamless and coordinated operations; with better information sharing and improved decision making.
Autonomous supply chains
By combining artificial intelligence with IoT, a software ‘agent’ can act on behalf of a stakeholder along the supply chain. Agents can be programmed to use algorithms to share data, predict outcomes and even negotiate with other agents, so they can pre-empt disruptions with timely interactions such as adjusting order quantities and lead times.
For example, agents can optimise truck loads — even at short notice — saving energy and making operations more sustainable and resilient. The hassle of integrating another IT system is also removed as agents can be added onto existing systems.
Autonomous supply chains can therefore help reduce labour costs and enhance operating efficiency by automating routine tasks of supply chain monitoring. By adjusting orders and learning to negotiate between different parties in the supply chain, autonomous supply chains can connect information both up- and downstream.
Another benefit of adopting this approach is that it helps open up the market to everyone, from small- and medium-sized companies to large multinationals. The value gained, thanks to better efficiency, is shared across the whole supply chain.
Are autonomous supply chains feasible?
Although some manufacturers have invested in developing the technology, there are currently very few industrial adopters of autonomous supply chains. Whether or not the manufacturing industry can integrate digital technologies and establish autonomy throughout the supply chain remains to be seen.
Working with local Cambridge tech company Fetch.ai to assess the technical feasibility of autonomous supply chains in real-life, our team from the Supply Chain AI Lab at the IfM has developed two autonomous supply chain demonstrators. These present feasible and scalable solutions by showing how IoT, multiagent systems and AI can work together to create autonomous, decentralised operations in supply chains.
With Fetch.ai, we were able to use the company’s new decentralised Autonomous Agent Framework (AEA) and corresponding technical support to implement a prototype using their agent development technology.
Using a mock scenario which mimics the supply chain of a meat company which purchases meat wholesale and supplies to local restaurants, we wanted to solve the need for rapid, secure exchange of data over the end-to-end of the supply chain.
This first demonstrator showcases how sensor data can be shared between organisational boundaries on a food logistics chain so that product quality can be guaranteed and traced. Some of the capabilities of the system include: the automatic selection of suppliers, the use of IoT in monitoring the ambient conditions of the transport vehicles, rerouting when unforeseen events occur (traffic jams, for example), and the provision of an analytic summary about product quality.Building on this, we developed a second demonstrator which streamlined the supply chain automation pipeline and integrated an agent-based autonomous supply chain management platform, including procurement, transport monitoring, negotiation, inventory update, and product quality summary.
The two demonstrators showcase how agent-based technology can integrate with IoT, AI, visualisation and web interfaces to form a prototype system for automating many mundane tasks in supply chains. They offer intuitive and tangible interpretations of key autonomous supply chain components to industrial stakeholders.
New opportunities for business
Autonomous supply chains offer new opportunities for businesses, helping them to understand their success in improving performance and helping supply chains act with immediacy and decisiveness.
We believe autonomous supply chains could be transformational for businesses, helping them to create better visibility and traceability, and automating routine operations without investing in integrated, expensive platforms.
Although integration with existing systems remain a barrier to adoption, robust platforms are emerging and have the possibility to make agent-based supply chains a reality.
Find out more about the IfM’s Supply Chain AI Lab: https://www.ifm.eng.cam.ac.uk/research/supply-chain-ai-lab/
