Article Preview
Top1. Introduction
The current structure and composition of our worldwide transportation system is extending exponentially and is a source of costs for all actors involved in it, should it be for the state that builds and maintains the roads and the road signs, or for the vehicle owners who buy expensive cars equipped with a lot of sophisticated devices. In that ecosystem, there is an emergent and urgent need to reduce the emission of carbon dioxide gas, to decrease the growing of traffic jam (Aridor et al, 2000), to limit the over-abundance of traffic signs, and to improve the interoperability of traffic sign between different countries (e.g. Japanese signs are often not understandable for European people), etc.
Aside from the technology exploited for driving (road marking, road signs, traffic rules, GPS, etc.) in the Existing Mobility Paradigm (EMP), new technologies also arise and offer new opportunities for the forthcoming years. Three of them are especially worth considering as they consist of the three concepts that support the funding of our paper. The first one is the high performance computing (HPC) (Byrnes et al., 2016) that allows aggregating computing power to generate exponentially more powerful systems able to solve large equations using deep learning techniques. The second one is the resolution of the geo-positioning satellite, which is already around a few centimeters1. This second element is to put in parallel with the development of street maps related to the infrastructure such as the one provided by Open Street Map (OSM) (Ly et al., 2015). Finally, the last arising technology is the 5G technology that allows a big and fast exchange of information over time (Mitra et al., 2015). All three technologies putted together constitute the pillars of a Future Mobility Paradigm (FMP), an arising new driving paradigm
This new paradigm aims at organizing the traffic based on a global coordination of the driving parameters with the ineluctable support of the three new technologies. FMP uses parameters including constraints associated to roads segment or to legislation, but also features associated to vehicles, to drivers, and to the context (like the weather or the time of the day). It has the ambition to leverage the protection of the environment and of the people, and in the meantime to reduce the cost associated with the movement of the goods and the people. Considering this new paradigm, we propose in this paper a traffic management algorithm based (1) on the clustering of vehicles having one or more common feature(s), (2) on computing the optimal and the global movement of the vehicles and (3) on automatically driving the cluster from a point A to a point B, potentially, without the intervention of human.
This paper suggests a new approach for the future of mobility. It is based on observations of the existing situation and technologies and on common sense. FMP has recently been proposed and the cluster based mobility disclosed in a patent (Feltus, 2018).
The paper is structured as such: first we present the expected Future Mobility Paradigm, then we propose a new cluster based model running in this new paradigm, thirdly, based on that model, we introduce a four-steps cluster based mobility deployment. Afterwards we give an example of usage of this deployment, we present the first steps toward the deployment of the new paradigm and we conclude the paper.