Traffic routing in urban environments: The impact of partial information

Abstract

Abstract-There are many studies concerning the problem of traffic congestion in cities. One of the best accepted solutions to relieving congestion involves optimization of resources already available, by means of balancing traffic flows to minimize travel delays. To achieve this optimization, it is necessary to collect and process Floating Car Data (FCD) from vehicles. In this paper we evaluate the repercussions of partial information on the overall traffic view, and consequently on the outcome of the optimization. Our study focuses on the role of the user participation rate and the availability of Road Side Units to collect the FCD. By means of simulation we quantify the impact of partially-available information on the computation of route optimization, and how it impedes traffic flows. Our results show that even minor uncertainties can significantly impact routing strategies and lead to deterioration in the overall traffic situation.

Deep Learning System for Vehicular Re-Routing and Congestion Avoidance

“…After reviewing previous works, we observed that one of the most used scenarios to evaluate proposals is the scenario known as Manhattan (or grid type) [25][26][27]. In our case, two synthetic networks were designed based on the Manhattan scheme: one of the 4 × 4 (M4) and another 5 × 5 (M5).…”

Partial and Local Knowledge for Global Efficiency of Urban Vehicular Traffic

“…In [5], the authors use partial knowledge of the network learned by users in a centralized context -a server receives voice tweets from drivers on traffic events. In [6], they quantify the impact of partially-available information on the computation of route optimization.…”