SIMCenter & DENSO give you your morning back

If you live in a congested city like Dublin, Ohio, you would be hard-pressed to find a ‘perfect’ commute. DENSO, in collaboration with the City of Dublin, is looking to give you your morning back. With data logged from cameras and radars installed at key intersections and traffic signals, SIMCenter researchers Punit Tulpule and Jagruti Patil are grasping the current traffic issues that plague Dublin drivers.  

The duo is developing algorithms using Python, Matlab, and ASN.1 OSS Nokalva to retrieve traffic flow rates, trajectories of vehicles, uncommon events, and queue lengths at the traffic intersections. Using a custom-trained ‘You Only Look Once (YOLO)’ algorithm, researchers will receive processed video data analytics. This data is used to count vehicles and pedestrians by class and obtain their trajectory at an intersection. Then, the data is fused with the Signal Phase and Timing to obtain the number of vehicles waiting at a signal. Punit and Jagruti will use sensor fusion methods to integrate the video data, basic safety messages from connected vehicles, RADAR data and personal safety messages to paint a complete picture of a driver’s commute, highlighting opportunities to save time and money with an emphasis on safety.  

For the second phase, SIMCenter researchers conducted a survey of Dublin residents to understand their commuting experience. During upcoming discussions with the traffic engineers of the City of Dublin, Punit and Jagruti will propose several solutions— the pre- and post-deployment data will highlight exactly how these solutions will cut down on commute times and save money for both the city and drivers. The project is set to end in December 2020.  

If you are interested in learning more about their work or would like to discuss potential research, Contact Punit Tulpule at tulpule.3@osu.edu, and Jagruti Patil at patil.198@osu.edu.