Development of an Application Using a Clustering Algorithm for Definition of Collective Transportation Routes and Times

This article presents the modeling, development and theoretical grounding for the development of an application based on the clustering algorithm DBSCAN, aiming to reduce the daily waste of time on the locomotion of a huge number of people to a common place. The clusters are created based on attributes, like the departure time of each person from its residence, the final destine and its both geographical locations (departure and arrival). People that compose the cluster are transported in a vehicle allocated according to the size of the cluster. A case study, with a specific group of people going to CEFET-MG Campus II, was conduct using a popular traffic simulator in order to measure the individual and the global time that people need to move, using the new transport arrangement. The simulation results were compared with a scenario where all people use public transportation. This comparison identified an average reduction of 45.80% in the time that people spent daily for their locomotion.