Supply chains today are a complex and interconnected system with emergent, often invisible dependencies. Such lack of visibility has hindered effective risk planning, resulting in late discovery of problems such as contamination of products, unsustainable manufacturing practices or reliance on suppliers located in risky geographical areas. Traditional initiatives that rely on manual data collection often fail due to supply chain complexity and suppliers’ unwillingness to share data. The Manufacturing Analytics Group at Cambridge has been developing supply chain risk surveillance techniques that do not rely on manual declaration of dependencies. We perform proactive monitoring of digital data to allow firms to infer hidden dependencies in the supply chain using machine learning. We are working on a neuro-symbolic framework based on knowledge graph and graph neural networks to reason over hidden links in the network. In this talk, we will discuss our approach and research directions.
About the speaker
Edward Elson Kosasih is a Ph.D. student at the University of Cambridge, working with Dr Alexandra Brintrup. His research focuses on applied machine learning for supply chains and manufacturing. He works closely with Aviva and Cambridge Centre for Data-Driven Discovery for his Ph.D.
This event will be a HYBRID event. To join this seminar online or in person, please register on Zoom. You will then receive an email with the dial in details for those attending virtually. Please check your spam/junk folders for joining details.
Contact firstname.lastname@example.org for more information.