This paper proposes a novel, nicknamed the "Pinpoint", method for an efficient and robust traffic estimation system using mobile phones as traffic probes. In this approach, the real-time traffic data is collected and sent to the server at the right time by the right players by which only the utilized data is reported by travelling vehicles. Mobile phones from walkers are prevented from sending data to reduce the data transmission load. In additions, this approach consists of a robust vehicle classification method based on only the GPS data. This novel feature improves not only the accuracy in estimating the seriousness of congestions but also the scalability of the system. This proposed approach can be flexibly applied in any traffic system structure, especially in developing countries where a lot of motorbikes are travelling on the roads. The evaluation shows that our proposed method is more efficient, effective and scalable compared to the existing ones.