Analysis and Improvement of Employee Transportation System

Abstract

This study involves in analysis and improvement of the employee transportation system in a private university in Izmir, Turkey. The problem is to design a transportation system using systematic and scientific methods. The input of the problem is the location of home addresses of the employees. It is required to determine the number and types of vehicles to be used to transport the employees to and from the university campus. Moreover, the routes of the vehicles should also be defined when the problem is solved. A two-phase solution approach is adopted. In the first phase, the addresses are partitioned into clusters by using K-means algorithm. The centres of clusters would be the bus stops in the transportation system. Once the cluster analysis delivers the cluster centres (location of bus stops) and the number of employees assigned to each bus stop, in the next phase, the school bus routing model (SBRP) is employed to solve the remaining part of the problem. In this phase, the number and types of the vehicles are determined as well as the routes of the vehicles. The details of clustering process discussed, and solutions of math models are shown for large data sets.