Analysing the Evolution of Kenya’s Road Network since the 1950s using Historical and Contemporary Road Datasets
Keywords: Historical Roads, Sub-Saharan Africa, Topographical Maps, Deep Learning, Infrastructure, Development
Abstract. Kenya’s road network has undergone substantial but uneven transformation since the 1950s. This study compares historical road data derived from topographic maps using a deep learning-based extraction method with contemporary datasets from the Kenya Roads Board, in order to trace long-term development patterns and regional disparities. To evaluate national progress and identify persistent structural gaps, changes in total road length and shifts in the average distance to the nearest road are calculated as primary indicators. Additionally, road expansion has been linked to population growth showing where infrastructure has lagged behind demographic change and where investments have been disproportionately high. A regional analysis of selected areas in southwestern Kenya between the 1950s and 1980s further illustrates contrasting development trajectories shaped by political priorities and settlement patterns. The results show that the southwestern, central, and coastal regions benefited from greater investment, while arid and sparsely populated northern counties remained underserved. Despite major improvements, spatial inequalities persist and rapidly growing urban centres, such as Nairobi, face challenges as population growth outpaces road expansion. It can therefore be concluded that Kenya’s road development has not followed the principle of spatial equity, emphasising the need for targeted interventions to address historical marginalisation and the demographic pressures present in rapidly growing urban regions.
