Simulating mobility to plan for electric minibus taxis in Sub-Saharan Africa's paratransit

Sub-Saharan Africa (SSA) is faced with the challenge to integrate e-mobility into its paratransit (its informal mass transit). Old, unsafe, fuel-inefficient, polluting minibus taxis are the cornerstone of daily commuting for millions in the region. Planning for electrification requires accurate high-frequency mobility data, which is currently unavailable. We analyse and improve on existing simulation models, which predict the energy usage with a micro-traffic simulation, SUMO, that up-samples low-frequency mobility data. We show that, compared to using measured mobility data, the current simulation approach overestimates energy expenditure. Results show a mean energy expenditure per distance overestimation of 14%, and mean energy per trip overestimation of 46%. We identify and virtualisation errors in the current driver and infrastructure models, and identify shortcomings of imposing a virtual road network with simulation software in the SSA context. We recommend and demonstrate virtualisation improvements for accurate electro-mobility planning in future.