REMIND - Reverse Manufacturing Innovation Decision Systems

This work investigates the operations of a logistic and waste recycling center where materials are collected by the company trucks fleet and then separated in order to be converted in secondary raw materials. The activity is characterized by low margins, uncertainties in supplies, and difficulties to track both materials and trucks flows. In these settings, the aim of our work is to develop three models that would be working in an integrated manner. Indeed, just like in the circular economy closed loop where all waste should feed another process, in the presented framework each model output become an input of the following model. A first mixed integer programming model is used to schedule and lower the selection operations of a two-phase waste selection process, while a pick-up and delivery routing model supports the logistic operations. In the end, a machine learning model performs customer profitability analysis once both logistic and selection costs have been first optimized and then allocated to each served customer.