A breakthrough solution for real-time counting stations has been successfully tested and is now available across the US, from mobility data providers Cellint. Tests of the TrafficSense system compared results from its virtual counting stations to real-time measurements from physical sensors, and showed less than 6% absolute average difference.
The solution, which relies on anonymous data from the cellular network, monitors the entire network population at the mobile switching centres, which are integral to the network. All phones on the network are monitored anonymously in real-time, 24/7, regardless if the phone has an active GPS or not, and without any dependency on sporadic locations from mobile applications.
It is the only real-time vehicle counting solution in the world that does not require field installation and maintenance. Data can be provided to road operators and planners through live dashboard and real time XML feed, as well as through CSV files.
Historical information can also be generated from archived network data. Daily, weekly, monthly and annual trends can be viewed and analyzed. Vehicle volume measurements comparison between Cellint’s Virtual Counting Stations and physical sensors in Illinois
“Our virtual counting stations can help transportation agencies and cities monitoring frequent volume changes over their entire road network at a reasonable cost without the need to deploy and maintain highly expensive physical sensors” says Ofer Avni, Cellint’s CEO. “This solution is especially helpful due to the latest changes in traffic patterns caused by the pandemic, as we can’t rely anymore on AADT [annual average daily traffic]measurements once every couple of years.”
Cellint’s technology takes cellular-based traffic detection to the next level by using ground-truth cellular signalling maps as a location reference, so each control message reported on the network is assigned an accurate street location in real-time.
The combination of huge data quantities with street location accuracy of each data point enables TrafficSense to detect all slowdowns in real-time, similar to road sensors, as well as provide non-biased information of the entire population for origin-destination and volume analysis.