Integration of Real Time, Fixed Sensor Data and Simulation/Development and Evaluation of an Advanced Traveler Information System Using Vehicle-to-Vehicle Communication Systems


Hoe Kyoung Kim, Ph.D. Student, Georgia Tech      
Dr. Michael Hunter, Associate Professor, Georgia Tech
Dr. Randall Guensler, Professor, Georgia Tech
Dr. Richard Fujimoto, Professor, Georgia Tech

Project Overview:

With the recent advent of more advanced technologies, particularly wireless communication technology, more economical traffic information systems are becoming possible.  Numerous researchers are exploring in- and between-vehicle communication systems, focusing on hardware and software development and testing, aiming to achieve more efficient and effective exchange of information.  Given these recent advances in information technology, the opportunity now exists to evaluate the utilization of dynamic, vehicle-oriented information exchanges with respect to their impact on transportation system performance. 

This research project is targeting the improvement of arterial road performance through the integration of real time fixed sensor data and simulation.  This project has also examined the exciting possibility of developing and evaluating an advanced traveler information system (ATIS) using vehicle-to-vehicle (V2V) communication systems with a microscopic simulation model.   

The V2V model was tested on notional small traffic networks (non-signalized and signalized) and a 6 x 6 typical urban grid network (signalized traffic network).  The V2V model consisted of three key modules: vehicle communication, on-board travel time database management, and a Dynamic Route Guidance System (DRGS).  In addition, the V2V system performance was enhanced by applying three complementary functions: Autonomous Automatic Incident Detection (AAID), a minimum sample size algorithm, and a simple driver behavior model.  To select appropriate parameter ranges for the complementary functions a sensitivity analysis was conducted.  The model performance examined network performance characteristics relative to three underlying system parameters: traffic flow, communication radio range, and the penetration ratio of participating vehicles in the V2V system.  In addition, V2V approach was compared to similar network capabilities only using a centralized traffic information system

This research found that an enhanced V2V model outperforms the basic model in terms of travel time savings and produces more consistent and robust system output under non-recurrent traffic states (i.e., traffic incident) in a simple traffic network.  Traffic incident detection time and drivers’ route choice decision rules are the most crucial factors influencing system performance.  As traffic flow and penetration ratios increased, the V2V system becomes more efficient, with non-participating vehicles also benefiting from the re-routing of vehicles that are sharing link performance information.  Different communication radio ranges were not found to have a significant influence on system operations.  It was found that a dynamic infrastructure-based traffic information system could replace a fixed infrastructure-based traffic information system with enhanced benefits, allowing for considerable savings in fixed costs and ready expansion of the system off of the main network corridor.

Sponsored by National Science Foundation EFRI ARES-CI (0735991)