The socio-spatial dynamics of a city undergoes constant changes over time. Consequently,the road network and the public transport system need continuous optimization to meet citizen demands. An alternative to reduce costs and impacts on evaluation of solutions is the use of simulators and models consistent with reality. Considering that, we processed vehicle tracking data and bus system planning information of São Paulo to improve the bus movement model used by InterSCSimulator, a highly scalable simulator for smart cities. In this paper, we present a mobility model based on real data from the São Paulo bus service to make the simulator more effective when recreating urban mobility scenarios.