Intelligent Management Functionality for Improving Transportation Efficiency by Means of the Car Pooling Concept

Information and communication technologies (ICTs) have long been attracting research interest, which is reflected in the design and development of powerful and complex network infrastructures, advanced applications/services, efficient power management, and extensions in the business model. A field of applications where ICTs find prosperous ground is transportation, in the sense of utilizing ITC findings in developing intelligent transportation systems (ITSs), i.e., systems for increasing transportation efficiency. This is motivated by the fact that transportation is associated with several drawbacks, e. g., with regard to continuously increasing traffic congestion in large cities worldwide. A transportation alternative to address congestion is the concept of car pooling, i.e., sharing a vehicle toward a common destination, based on a priori agreements. The goal of this paper is to present novel management functionality for dynamic ride matching, within a car-pooling context. The functionality uses previous knowledge in proposing valid car-pooling matches. Knowledge is obtained through the exploitation of Bayesian networking concepts, specifically the Naive-based model. Simulation results showcase the effectiveness of the proposed functionality, the advantage of which lies in the fact that the reliability of the knowledge-based selection decisions is higher. This means that there is higher probability of satisfying the drivers' and passengers' preferences through the selected matches.