Collective Supply Chain Resilience (CORES)

Public description

Supply chain resilience is a widely studied topic of significant impact on our society. As organisations outsource production to one another they create economies of scale and reduce prices but also increase the risk of disruption that cascades throughout the entire supply chain if any member of the chain is disrupted. Typically, organisations act alone, rather than as an ecosystem when predicting disruptions and deciding on mitigation strategies. However, disruption data an individual organization can collect and analyse is small, imbalanced, and partial entirely to its own view. When uncertainties increase, this individualistic approach results in short-sighted decisions. Numerous studies proved that _increased data sharing_ and _collective decision-making_ would _increase resilience_, but this has not been plausible as members of the ecosystem fear that information shared can be used opportunistically by other parties.

_Federated learning_ is an emerging approach in the Artificial Intelligence field that may help supply chain members collectively optimise resilience while keeping their data private. The approach enables organizational agents to collaboratively develop a shared prediction model. Here, if one organization is able to predict a disruption, its knowledge can be shared sending an early warning signal and giving companies time to respond. As the approach can be automated, costs of manual orchestration are avoided.

In this project, we will develop, validate, and compare suitable federated learning models specifically for disruption prediction and collective learning in supply chains with real use cases in the aerospace and manufacturing sectors.