Formation of traffic jams. Modeling and communication technologies

In an environment where the car is still the preferred means of travel, and given the renewed need for mobility, understanding road congestion has become a major challenge for modern societies in terms of lost productivity, lost time, greenhouse gas emissions and other negative externalities.

Traffic jams have numerous economic, social, logistical, health and ecological consequences. In August 2010, a 100-kilometer traffic jam in Beijing lasted 12 days, caused by trucks bringing in materials for work on the G110 freeway. A study by the research institute Centre for Economics and Business Research and Inrix, an American traffic information company, published on Tuesday December 17, 2013, shows that traffic jams cost the French economy 5.9 billion euros, every year (reference: https://cebr.com/reports/the-future-economic-and-environmental-costs-of-gridlock/ ). The direct costs of congestion (value of fuel and time lost) for all Paris households are expected to rise from $6.2 billion to $10 billion between 2013 and 2030 (CEBR, 2014). In addition to reducing traffic speed, congestion is a major source of air pollution.

The efficiency of a country's transport infrastructure and systems is a major factor in its economy. On this chessboard, road transport continues to play a predominant role. Since the Thirty Glorious Twenties, with the intensification of trade, the reason for travel, and the multiplication of the number of vehicles, the vector of travel, infrastructures have always had to be adapted to the volume of traffic. Today, this policy of network expansion is no longer economically tenable or socially acceptable. Indeed, in a constrained global environment, collective awareness of the negative impacts of such an approach is leading to a paradigm shift. The use of today's congested infrastructures must therefore be optimized to provide users with a higher level of service.

The appearance of congestion is a phenomenon that has been studied for over fifty years, and the mechanisms at play on homogeneous sections are relatively well understood. This is somewhat less true of the global approach to a network made up of homogeneous sections and discontinuities. What is also less well understood is how to intervene effectively on this system in order to get the best out of it. However, we need to qualify our ambitions, because in reality, the idea of smooth traffic flow does not exist per se. The fluidity of traffic leads to changes in the choices made by agents, which can ultimately lead to congestion. A problem akin to Jevons' paradox, or the "rebound effect".

The direct costs...