A novel approach to assess a street network and find routes using static and dynamic data with AHP and ACO methods

Abstract

Vehicular traffic has been increasing due to the natural growing of the cities, and the immigration of people from rural areas to big cities. Thus, everyday it is more critical how traffic conglomerates on the main arterials of the cities; even if they are upgraded by building bridges, new shortcuts or streets, these adaptations are temporal solutions because traffic increases faster than the modifications, and finally these actions are insufficient as permanent solutions to this problem. An alternative solution is to take the better profit of the available infrastructure in order to alleviate this problem as much as possible. In such conditions, the solution has to consider not only the static characteristics of the streets and avenues, as number of lanes, mean speed, and sense to mention some of the most important, but also the dynamic conditions that can affect the traffic from slight to strong. Many algorithmic approaches have emerged to tackle this problem by considering space and time in a combinatorial way to suggest possible routes, as the fastest or shortest within the origin-destination search space set, even some begin to include crowdsourcing for better results. An alternative solution is the main contribution of this project, where we present the development of a methodology for a street network assessment and routes finding, under static and dynamic constraints, that offers the possibility for drivers to identify the current situation of the street segments by their status to provide their service because of theirs characteristics and find the most logical options to travel from an origin to a destination, but also to find other interesting options to drive through avoiding bottlenecks, due to the rush hours, accidents, or any other problem. This methodology is based on the AHP method, which allows us to generate a simplified map from the original one, by assessing the network, and taking into account the static and dynamic aspects that affects traffic, the latter gathered from a crowdsourcing and a smart and digital city approach, and that depends on the criteria, as time or distance, selected by the user; the methodology also integrates an ACO algorithm, that uses the simplified map to find a solution in a reduced time, as the mean to find the more attractive routes for a particular journey. The results show the advantages and how promising is this proposal that can be used, not only for traffic problems, but also for city planning, as different scenarios can be easily displayed in a simplified map, and the static infrastructure can be better planned as the city is growing.