Spatio-temporal optimal law enforcement using Stackelberg games.

Abstract

Every year, road accidents claim the lives of around 1.2 million worldwide (USDOT-NHTSA, 2012). Deploying speed traps helps bounding vehicles speed and reducing collisions. Nevertheless, deterministic speed traps deployment in both spatial and temporal domains, allow drivers to learn and anticipate covered areas. In this paper, we present a novel framework that provides randomized speed traps deployment schedule. It uses game theory in order to model drivers and law enforcers’ behavior. In this context, Stackelberg security game is used to derive best strategies to deploy. The game optimal solution maximizes law enforcer utility. This research work aims to optimize the deployment of speed traps on Lebanese highways according to the accidents probability input data. This work complements the near real time accident map provided by the Lebanese National Council for Scientific Research and designs an optimal speed trap map targeting Lebanese highways.