Connected Vehicle Pilots Project Adapts Existing Software for Local Deployment

Tampa Hillsborough Expressway Authority (THEA), along with New York City Department of Transportation and the State of Wyoming DOT, are currently in the installation stage of the Connected Vehicle (CV) Pilot, a ground-breaking project of multiple Connected Vehicle applications that will help improve safety, mobility, and lessen environmental impact. In Tampa, the project involves installing radios and computers in over 1600 cars, 10 buses and streetcars, as well as roadside computers at over 40 fixed locations in downtown intersections. These radios and roadside units will enable vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication in the Central Business District (CBD). The project is partly funded by the U.S. Department of Transportation (USDOT), and partly by THEA.

All of the connected vehicle applications being implemented and evaluated in this pilot program did not need to be developed from scratch. Software developers for the three CV Pilot sites have been able to take advantage of software prototyped by many programs, including the USDOT’s Dynamic Mobility Applications (DMA) program and the Safety Pilot Model Demonstration project in Ann Arbor, MI. Prototype software developed for these programs is freely available as open source software on the USDOT’s Open Source Application Development Portal (OSADP) at http://itsforge.net.

In many cases, however, existing software has required modification or customization to be implemented in the CV Pilot locations. One of the major goals of the CV Pilot project has been to document the sites’ application development process in “Application Deployment Plan” reports, including adaption of existing software. This paper discusses two instances where THEA significantly modified existing CV applications to achieve THEA’s goals for the project.

One of THEA’s six use cases is to prevent the entry of vehicles going the wrong way (against the permitted flow of traffic forTraffic Signal the given time of day) onto the Reversible Express Lanes (REL) of the Lee Roy Selmon Expressway. Entry into the REL is governed by a traffic signal (see photo ) so THEA started by using existing Red Light Violation Warning (RLVW) software. The plan was to treat the traffic signal as if it changed only twice per day (i.e., at the times the directionality of the REL was reversed) and that a vehicle trying to enter the REL the wrong way could be detected as if it were trying to run a red light.

CVP MapHowever, the complexity of the intersection (cars could enter the REL from three intersection approach directions, and could go in legal directions as well as the illegal direction) made this approach unworkable. THEA evolved the RLVW application into a Wrong Way Entry (WWE) application. The key change was to adapt the Signal Phase and Timing (SPaT) and MAP (intersection geometry) messages broadcast by the traffic signal to include a bit for each approach lane indicating whether the direction for that lane is “revocable”, i.e. changeable. By examining the “revocable” bit for the lane of an approaching vehicle and the time of day, the vehicle’s Onboard Unit (OBU) can determine whether it is going the wrong way. If so, it will send a warning message to the vehicle’s driver and broadcast a warning message to the drivers of approaching vehicles.

The second application that required adaption was the Pedestrian Crossing (PED-X) application, originally planned to warn pedestrians crossing Twiggs Street of approaching vehicles that might be on a collision course. The mid-block pedestrian crossing connects the county courthouse and the nearest parking garage. A Roadside Unit (RSU) will convert Basic Safety Messages (BSMs) broadcast by approaching vehicles to WiFi messages. Mobile devices (smartphones) carried by pedestrians will receive these messages and compute the probability of a collision, based on the pedestrian’s current location and walking speed. However, the mobile device cannot determine the pedestrian’s location and speed with sufficient accuracy - for example to distinguish stepping into the street from standing on the sidewalk - to avoid numerous false alarms. Therefore, the PED-X application will save generated warnings to the device for future study and analysis, but will not issue any warnings to the pedestrian. The mobile devices will also generate, save, and transmit via WiFi Personal Safety Messages (PSMs) similar to BSMs, but these PSMs will not be used to generate alarms to approaching vehicles.Pedestrian Crossing

THEA will modify the vehicular side of PED-X to create the Pedestrian Collision Warning (PCW) application. Since the PSMs from mobile devices do not provide sufficient accuracy for reliable warnings, the PCW application will use pedestrian locations from two Lidar sensors installed near the crosswalk. The Lidar sensors will send pedestrian location information to the RSU via landline. The RSU will generate PSMs and broadcast them via Dedicated Short Range Communication (DSRC) to the OBUs of approaching instrumented vehicles. Each instrumented vehicle’s OBU will determine whether a collision is possible, and if so will issue a warning to the vehicle’s driver.

Additional information about THEA’s Connected Vehicle deployment may be found at https://www.tampacvpilot.com/ and https://www.its.dot.gov/pilots/.