Extracting kinematic data for calibration of microsimulation models using smartphone probes

This study introduces a new methodology for acquiring vehicle tracking data with which to calibrate and validate microsimulation traffic models. Most common approaches are based on video image processing algorithms or the use of roadside bluetooth detectors. In this paper a procedure is presented that makes use of GPS-equipped smartphone probes supplemented by ground triangulation. This depends on the capability of assisted-GPS to capture individual vehicle movements over time for changing road and traffic conditions, where satellite signal strength is subject to interruption with corresponding errors in the estimated position of the receiver. This is a major concern in dynamic traffic conditions where signal strength can vary appreciably depending on roadside land development densities, signal occlusions caused by other vehicles in the traffic stream and adverse weather conditions. When probably adjusted for error, vehicle position information and speeds obtained from smartphone probes can be instrumental in identifying factors that affect road safety at a given location. Using the calibrated simulation platform, they can also be used to evaluate the safety implications of different countermeasures introduced at a given location. Current ITS allow the acquirement of traffic data from fixed sensors, but fail to accurately track vehicles and to produce detailed information about vehicles interactions. With a significant penetration rate, smartphone probes can provide representational coverage of traffic flow conditions over time. In this study, a microscopic simulation model was calibrated based on kinematic data acquired from a sample of smartphone probes deployed along three test road segments with different land development densities. The calibrated model was then applied to a case study two-lane highway in southern Italy (SS106), and appropriate measures of safety performance were estimated. A number of indicators were used to reflect safety performance, such as, differential operating speeds, acceleration profiles, speed variance, etc. For each simulated indicator, potential conflict scenarios were evaluated in terms of number of vehicles in conflict and exposure time to conflict. Safety performance was expressed in terms of indicators representing vehicle interactions for different vehicle pairs in the traffic stream. These safety performance measures were compared to the restrictions imposed by road geometry along the case study highway. Preliminary results suggest that GPS-assisted smartphone probes are useful for providing meaningful experimental traffic for calibrating simulation models and potential road safety issues.