Publication Date

Summer 2016


School of Engineering and Computational Sciences


Engineering: Industrial and Systems


logistics, mTSP, k-means, genetic algorithm


Operations Research, Systems Engineering and Industrial Engineering


The design of the distribution process is a strategic issue for almost every company. As the use of advanced technology and automation increases in manufacturing and logistics, the implementation of autonomous and electrical transportation, such as driverless vehicles and electric trucks, has become an interesting topic of study within the last few years, with the main objective of minimizing distribution costs and delivery times. The purpose of this research is to prove that intermodal delivery networks, which may combine a train and several electric vehicles, are more efficient and environmentally friendly than unimodal networks for high volume and long haul transportation, regardless of the customers’ distribution. This is only applicable if demand does not fall within the capacity restriction of road transportation vehicles. To do so, this paper utilizes an optimization algorithm that consists of a feedback mechanism between K-means and a genetic algorithm, which finds the optimal routes between distribution centers and surrounding customers as a multiple traveling salesman problem (mTSP).