An event-driven optimization framework for dynamic vehicle routing

The real-time operation of a fleet of vehicles introduces challenging optimization problems researches in a wide range of applications, thus, it is appealing to both academia and practitioners in industry. In this work we focus on dynamic vehicle routing problems and present an event-driven framework that can anticipate unknown changes in the problem information. The proposed framework is intrinsically parallelized to take advantage of modern multi-core and multi-threaded computing architectures. It is also designed to be easily embeddable in decision support systems that cope with a wide range of contexts and side constraints. We illustrate the flexibility of the framework by showing how it can be adapted to tackle the dynamic vehicle routing problem with stochastic demands. Computational results show that while our approach is competitive against state-of-the art algorithms, it still ensures greater reactivity and requires less assumptions (e.g., demand distributions)