Application of Bluetooth Technology for Mode- Specific Travel Time Estimation on Arterial Roads

The tendency to use Bluetooth Technology (BT) for travel time estimation is increasing due to growing number of Bluetooth-enabled devices among road users, anonymity of BT detections, flexibility of deployment and maintenance of Bluetooth sensors etc. Although Bluetooth has been demonstrated as a promising technology, there remain problems which affect the accuracy of the estimation such as difficulty of distinguishing between multiple travel modes (e.g. motor vehicles, bicycles and pedestrians). This study aims to examine the feasibility of estimating mode-specific travel time using BT data under uncongested traffic conditions. In this context, three clustering methods Hierarchical, K-means and Two-Step are used as the core techniques for classification. The results show that the methods can successfully distinguish between motor vehicles and bicycles from BT detection events, resulting in accurate travel time estimation for motor vehicles.