The goal in Robust Optimization is to optimize not only the quality of the solutions but also the variation of this quality with the uncertain parameters of the optimization problem. We propose a robust model for the bi-objective shortest path problem applied in a smart mobility context: Finding routes for cars in a city to minimize travel time and gas emissions. Our proposal treats robustness from a multi-objective point of view. We model the parameters that define each instance as random variables, described through their mean and variance. In this way, we can obtain efficient solutions that are also less sensitive to changes in the environment. We run different types of algorithms in multiple instances to solve this problem so that we obtain a global view of the behavior of different techniques. All experimentation uses a scenario based on real data: The province of Malaga, Spain. This realistic settlement for our study allows us to test the applicability of our model in final systems for the citizens.
The results clearly state the interest of our proposal for tackling robustness and represents a new state-of-the-art in smart mobility, an always appealing feature of works, that could lead to an industrial prototype.
