Utilizing cellular-based vehicle probing (CVP) data in traffic management the experience of deployment, testing and validation

Cellular-based vehicle probing (CVP) technology have gradually become the most potential source of traffic data collecting with the advance of cloud computing. Traffic data collection is the base of travel demand management. In the past, traffic data, such as flow, travel time and velocities of road segments, are obtained from the specific equipment like vehicle detector (VD), global position system (GPS) and electronic toll collection (ETC) system. These methods all need to invest huge amount of capital to build and maintain the infrastructure for the purpose of data collection. Compared to the other approaches, CVP may be a more cost-effective method for traffic data collection. In addition to travel time, this study proposed a model to estimate traffic flow and turning ratio, which are the most challenge parts, from CVP data. The result shows that the model has good estimation capability. The MAPE (mean absolute percentage error) of traffic flow and travel time are 20% and 16% separately.