This work presents, for the first time in literature, a low-cost framework to mine data obtained from passengers smart cards, buses GPS and bus stops geolocation using Lambda Architecture approach. Operators, companies, government and passengers will use this knowledge for improving usability, comfort, and quality of transportation service. This analysis gives greater insight into the volume and flow of passengers and the real existing demand for bus services, facilitating its control and management, allowing decision-making. As result, bus stops and city areas are classified according to buses demand.