Estimating online vacancies in real-time road traffic monitoring with traffic sensor data stream

Real-time road traffic monitoring is widely considered to be a promising traffic management approach in urban environments. In the smart cities scenario, traffic trajectory sensor data streams are constantly produced in real time from probe vehicles, which include taxis and buses. By exploiting the mass sensor data streams, we can effectively predict and prevent traffic jams in a timely manner. However, there are two urgent challenges to processing the massive amounts of continuously generated trajectory sensor data: (1) the inhomogeneous sparseness in both spatial and temporal dimensions that is introduced by probe vehicles moving at their own will, and (2) processing stream data in real time manner with low latency. In this study, we aim to ameliorate the aforementioned two issues. We propose an online approach to addresses the major defect of inhomogeneous sparseness, which focuses on employing only real-time data rather than mining historical data. Furthermore, we set up a real-time system to process trajectory data with low latency. Our tests are performed using field test data sets derived from taxis in an urban environment; the results show that our proposed method lends validity and efficiency advantages for tackling the sparseness, and our real-time system is viable for low latency applications such as trafficmonitoring.