Autonomous Vehicles and Braess Paradox
Belkina, Elizaveta & Zhestov, Egor & Shestakov, Artem. (2021). Methods for resolving the Braess paradox in the presence of autonomous vehicles. Computer Research and Modeling. 13. 281-294. 10.20537/2076-7633-2021-13-2-281-294.
http://www.mathnet.ru/php/archive.phtml?wshow=paper&jrnid=crm&paperid=884&option_lang=eng
While consumers and media outlets have been concerned about the safety of self-driving vehicles, have we truly thought about whether they would make our commutes faster? As autonomous vehicles become more prevalent on today’s roads, their role in the efficiency and performance of existing transportation networks must be considered. These concerns were addressed in a publication in the Computer Research and Model, where they discuss how the combination of common (human–operated) and autonomous vehicles on roads can cause Braess paradox by amplifying total network loads. In response to his dilemma, an algorithmic approach is taken to model roads where the amount of lanes of autonomous vehicles can travel is restricted. Through their computation, they suggest that assigning designated lanes to self-driving vechiles could eliminate Baress’ paradox.
Braess paradox is the phenomenon where adding additional resources to a transportation network can sometimes hurt performance at equilibrium. Specifically, in networks where an additional edge was added, the total cost of travel (time would increase across all of the shortest available routes. The total travel time is the Nash Equilibrium. While our lectures looked at Braess paradox with relation to Nash equilibrium, the researchers in this article instead utilize Wardrop equilibrium. Similar in concept to Nash Equilibrium, Wardrop shows that the travel time on all routes is not greater than the travel time of any vehicle traveling on unused roads, as drivers won’t choose routes with less traffic if the overall travel time is greater relative to other routes.