Speed estimation through mobile sensing

With the advances in wireless communication and mobile computing, a future infrastructureless self-organizing traffic information system, where vehicles can form a network for exchanging traffic information among themselves, will soon be realized. In an infrastructureless traffic information system, vehicles will act as mobile sensors and collect the traffic data as they travel. Smartphones are a great choice for traffic sensing devices as they are now equipped with a variety of sensors such as global positioning system (GPS) receiver, accelerometer, gyroscope, camera, and microphone. These sensors can be exploited to collect traffic data. Although there are many types of sensors available for traffic sensing, past studies have mainly focused on a GPS receiver. However, a GPS receiver consumes a lot of power and hence it can significantly shorten the battery life. In this paper, we explore the possibility of using other types of sensors on a smart phone for traffic sensing. Particularly, we investigate whether it is possible and how accurate it is to estimate vehicle speed from the data sensed by an accelerometer. The accuracy of our proposed estimation method will be presented.