As AI is gaining momentum in specific fields, supply chain problems and their inherent complexity remain important barriers in its widespread diffusion. Moreover, the lack of available data may prevent the necessary learning opportunity for algorithms to be developed.

As the Operational Research community used to do, also the AI research community use to resort to “challenges” as a way to overcome such hurdles, and achieve significant leap forward in theoretical and applied knowledge.

The goal of this thesis is twofold. First is to understand which problem in the SC design and management could work as suitable AI challenges, to become relevant fields for dedicated AI research. This will require to explore existing AI achievement, find gaps, analyze the generalizability of the possible problems so as to gain adequate attention from the AI research community. Once feasible problems have been found, the second goal is that of collecting adequate data and formal knowledge which would be needed by the contenders, thus providing the starting conditions for the challenge to happen