An automated pipeline routing system based on machine learning and mathematical programming

Abstract: Nowadays there is a large number of practical situations in which it is necessary to design networks optimally. The problem that motivates this project arises from the layout of pipelines in naval engineering, although it has direct applicability in other situations such as the use of intelligent infrastructures. In this project, we develop a methodology for the automated design of complex networks based on the decomposition of multi-commodity flow problems into separable sequential flows, avoiding the constructibility problems based on machine learning algorithms. This methodology will lead to a prototype that will be validated in the real environment of the collaborating agent.

Researchers: Justo Puerto Albandoz, Víctor Blanco Izquierdo, Yolanda Hinojosa Bergillos, Diego Ponce López, Miguel Angel Pozo Montaño, Gabriel González Domínguez

Partner: Andalucía Tech and Ghenova Ingeniería, S.L.U.

Related project: Un sistema automatizado de rutado de canalizaciones basado en aprendizaje automático y programación matemática

iMAT research line: RL4. Modeling Transport & Logistics