Data driven safe vehicle routing analytics: a differential evolution algorithm to reduce CO $$_{2}$$ 2 emissions and hazardous risks

Abstract

Contemporary vehicle routing requires ubiquitous computing and massive data in order to deal with the three aspects of transportation such as operations, planning and safety. Out of the three aspects, safety is the most vital and this study refers safety as the reduction of $$\hbox {CO}_{2}$$ emissions and hazardous risks. Hence, this paper presents a data driven multi-objective differential evolution (MODE) algorithm to solve the safe capacitated vehicle routing problems (CVRP) by minimizing the greenhouse gas emissions and hazardous risk. The proposed data driven MODE is tested using benchmark instances associated with real time data which have predefined load for each of the vehicle travelling on a specific route and the total capacity summed up from the customers cannot exceed the stated load. Pareto fronts are generated as the solution to this multi-objective problem. Computational results proved the viability of the data driven MODE algorithm to solve the multi-objective safe CVRP with a certain trade-off to achieve an efficient solution. Overall the study suggests 5 % increment in cost function is essential to reduce the risk factors. The major contributions of this paper are to develop a multi-objective model for a safe vehicle routing and propose a MODE algorithm that can handle structured and unstructured data to solve the safe capacitated vehicle routing problem.