E-GPS has been subject to robust testing at several toll and managed lane facilities in the US. At every site, test plans and application use cases were drafted by the agency. There were no hand-picked scenarios. Result: E-GPS met or exceeded expectations every time. The technology routinely and significantly exceeded the performance of video systems. It measurably functioned at at an equal-to-or-higher level of accuracy than the resident RFID solution. The results were the same on paint-separated and elevated express lanes. In short, E-GPS brings operator performance aspirations to life. Details on some of our pilots are below.
LA Metro Express Lanes, California
The Los Angeles County Metro system is one of the most challenging toll environments in the nation, consisting of elevated express lanes (EL) with general purpose (GP) traffic lanes below, and painted two-line separation of the EL adjacent to the GP lanes, causing lost revenue due to drivers bypassing the toll points.
LA Metro conducted a Proof of Concept (POC) to determine whether a system based on a location-aware mobile application with geo-fencing could capture toll transactions as reliably as the existing system of transponders, cameras and physical gantries. The results showed that GeoToll’s E-GPS could!
Test requirements were composed of twelve test cases including single point toll charging at LA Metro’s physical gantries, in-between gantries, on elevated roadways and on flyovers; multi-point toll charging which required the building of a trip; not detecting GP vehicles in the EL and vice versa; determining whether a vehicle illegally entered or exited an express lane; trip building with missed gantries; and confirming trip results independent of High Occupancy Vehicle (HOV) declaration; all tests were run with different smartphone brands. Tests were conducted in a similar manner as described in the BATA test (below) with qualified GeoToll testers driving a predetermined route on the I-110 Freeway.
The GeoToll system consistently performed above 99% accuracy during the first pass data analysis with over 10,000 virtual gantry/toll-point crossings. After re-applying GeoToll’s Supervised Machine Learning process to the larger data set than that used for calibration, E-GPS accuracy was shown statistically capable of achieving the sought-after "5-9’s" level.
Bay Area Toll Authority, MTC, California
The Bay Area Toll Authority pilot demonstrated the accuracy of GeoToll E-GPS in detecting toll collection points at the Bay Area toll bridges as well as the ease of use of the mobile payment application. A comparison of captured transactions using the FasTrak toll collection system, which includes transponders and license plate detecting camera enforcement was made with GeoToll’s E-GPS enabled phones in test vehicles.
Seven bridge locations where thoroughly tested with twelve use cases under different environments over two phases. Phase I tested two bridge locations at San Mateo-Hayward and Dumbarton - a prerequisite to do Phase II. Success was defined as the mobile app having a 98% or higher detection rate and no major issues that would adversely affect the customer experience. Test results showed that the GeoToll mobile app detected the test trips as recorded in the driver’s logs at a 100% success rate.
As a result of Phase I success, GeoToll was approved to run Phase II which consisted of testing at five additional bridges: Antioch; Benicia-Martinez; Carquinez; Richmond-San Rafael; San Francisco-Oakland Bay. Phase II followed the same methodology and success criteria of Phase I, with test returning comparable results: GeoToll detected the test trips as recorded in driver’s logs at a 100% success rate.
GeoToll has conducted 6 proof of concept test and is in the middle of 2 more. We are happy to discuss them upon request.