A markovian decision model on a personal computer is being developed for finland's roads administration to optimize the allocation of funding among rehabilitation actions, geographic regions, and traffic volume classes on roads with oil gravel and other light pavements. In the long term, the model recommends a funding allocation and condition distribution that minimizes the sum of agency and user costs; in the short term, it recommends a strategy to maximize annual progress toward the long-term goal. As traffic volume and network length grow, the system simulates their effect on rehabilitation policy.This paper appears in transportation research record no. 1291, Fifth international conference on low-volume roads, may 19-23, 1991, raleigh, north carolina, volume 1.
Abstract