Road authorities rely on accurate and timely road weather and surface condition information provided by road weather information systems (RWIS) to optimize winter maintenance operations and improve the safety and mobility of the traveling public. However, RWIS stations are costly to install and operate and therefore must be placed strategically to accurately monitor the entire highway network. Few guidelines are available for optimizing RWIS networks and thus maximizing return on investment.
This project developed several approaches for determining the optimal location and density of RWIS stations over a regional highway network. To optimize locations, three approaches were developed: surrogate measurebased, cost-benefitbased, and spatial inferencebased. The surrogate measurebased method prioritizes locations that have the highest exposure to severe weather and traffic. The cost-benefitbased method explicitly accounts for the potential benefits of an RWIS network in terms of reduced collisions and maintenance costs. The spatial inferencebased method maximizes the use of RWIS information to optimize the configuration of an RWIS network. To optimize network density, a cost-benefitbased method and a spatial inferencebased method were developed. To demonstrate the applications of the proposed approaches and evaluate existing RWIS networks, four case studies were conducted using data from one Canadian province (Ontario) and three US states (Minnesota, Iowa, and Utah).
It was found that all approaches can be conveniently implemented for real-world applications. The approaches provide alternative ways of incorporating key road weather, traffic, and maintenance factors to optimize the locations and density of RWIS stations in a region; the alternative to use can be decided based on the data and resources available.
End date: June 2016
Investigator: University of Waterloo