A study of the dynamic dispatch algorithm for a dial-a-cab taxi service

Abstract

A taxi dispatcher has to determine how to process customer demands as they arrive in real time in a way that most efficiently makes use of the resources. As the fleet size increases, human dispatching proves to be inconsistent and inefficient, therefore it would seem advantageous to make use of a machine algorithm to alleviate this problem. This project aims to prove the advantages of utilizing these machine algorithms over human dispatching through the utilization of accurate scheduling and optimal idle time placement. Two dispatch algorithms and an additional idle placement method are developed. A general service area and demand model is implemented to test the algorithms and compare their efficiency.