A Stochastic Route Choice Model in Optimal Control of Road Freight Flows: a Mexican Case Study

The inefficiencies arising from increased freight road traffic in the late 1990s in Mexico have posed to planners engaged in the road provision the goal to find new control measures to reduce the adverse impacts, such as road damage and overloading. In the 1990s, road freight traffic in Mexico altered notably after the deregulation of the road freight industry in 1989; the North American Free Trade Agreement (NAFTA) in 1994; and the privatisation of the Mexican Railway in 1997. These changes created new land freight flows, attracting attention to the characterisation of lorry flows on the roads, and the estimation of possible impacts in areas like: route choice, road damage and repair costs, overloading practices and government’s decisions as weight limits, tolls levels and loading enforcement scheme. In the Mexican case, there is a need for actions in freight traffic, and a first step is the yearly road survey promoted by the Mexican Transportation Secretariat since 1991, to gather freight lorry data on the paved road network (Rico, 1998). Interventions, however, need to be assessed before implementation, and this is the role for network modelling. The search for a realistic model based on the hauliers’ choices for routes suggests a stochastic focusing. This focus lets us represent hauliers’ choices, and their distinct perceived network attributes due to lack of perfect information, as well as coping with the problem of including in a model all the variables that could influence the utility in each choice (Kanafani, 1983).