Hurricane Proof of Concept: States’ Experience with Real-time Connected Vehicle Data
This pilot proof concept started as a basic question – “Could Connected Vehicle Data (CVD) be used as a real time data source to estimate volumes and if so, what would be needed to make this possible? So, The Eastern Transportation Coalition assembled a technical team and procured three months of Connected Vehicle Data along with access to an analytics platform to assess the viability of delivering and visualizing a real time CV data feed as well as devising a process that could validate the accuracy of the traffic volume estimates. The visualization of regional CV data as individual vehicles in a live data stream brings amazing intuitive value. CV data is growing in size and velocity and we are blazing new ground to visualize real time volume and O&D.
Six southeastern states participated in the pilot, hence the volume of data was huge. There were over 75B data points per month which came to over 230B data points in the 3-month pilot study period for the six states. It showed that managing CV data at scale is challenging for the industry and that sheer size and velocity of the data will require efficient processing architecture, as well as streamlined calibration, calculation and conflation techniques. For the validation process that is detailed in the report, simple scaling factors based on roadway class, time-of-day, day-of-week, and state were demonstrated to provide a workable and sufficiently accurate traffic volume estimate. Including additional factors, and leveraging more advanced machine learning may improve accuracy. Overall, the pilot clearly points to a viable path toward real-time traffic volume estimates based on CV data.
The Final Report is available here.