Prediction of gravel road roughness propagation using the MarKov process

Abstract

Gravel roads comprise a big percentage of the road network in low and middle income countries. Their surface roughness always causes excitation of vehicle suspension systems thereby influencing comfort and safety of road users. This research utilized surface roughness data for existing gravel roads to predict future surface roughness using the Markovian theory. As a consequence, optimization of resource allocation during planning and budgeting for future mechanized maintenance and rehabilitation of gravel roads can be done to decrease problems related to poor performance. The study objectives were: (i) to establish the current roughness condition of selected gravel roads in a study area, (ii) develop an approach for prediction of future gravel road roughness condition, and (iii) develop a resource allocation approach in utilization of resources for future maintenance and rehabilitation. Two roads were purposively sampled from amongst 17 major gravel roads in greater Kamwenge district. These were Kanara – Rweshama and Kahunge – Nkarakara – Kiziba roads. Roughness data were determined by conducting road distresses measurements (potholes, gullies, corrugations, rutting and stoniness) over four quarters covering the period January 2016 to January 2017. The results show that surface roughness existed at significant levels on the roads studied and propagated in space over the period of study with significant interaction between roughness and time. The Markovian theory based approach satisfactorily predicts future roughness based on the current state with at least 90% degree of accuracy though it is location specific. This approach can be adapted to other geographical locations. A resource allocation approach in utilization of resources for future maintenance and rehabilitation of deteriorated gravel roads has been developed in this study on the basis of projections of critical points and terminal serviceability in road life. This is helpful in planning and budgeting for resources required for future gravel road maintenance and rehabilitation. It is recommended that additional work is done covering different geographical and socio-economic parameters because the approach is climate, traffic loading and location material specific. The approach can also be used to inform policy for choices of road maintenance and rehabilitation by Local Governments. The developed approach is useful in resource allocation in the utilization of resources for future maintenance and rehabilitation of deteriorated gravel roads on the basis of projections of critical points and terminal serviceability points in road life.