Development of an Innovation Corridor Testbed for Shared Electric Connected and Automated Transportation
A number of testbeds are currently being developed across the United States to evaluate various applications of shared mobility, vehicle electrification, vehicle connectivity, and vehicle automation. This study utilized a testbed called the “Innovation Corridor” to establish a Connected Vehicle (CV) testbed locally in Riverside, California to support a variety of experiments. The testbed consists of a six-mile section of University Avenue between the University of California, Riverside campus and downtown Riverside. The testbed supports passenger vehicles, trucks, transit, bicycles, various forms of micromobility, and is continuously being instrumented with various infrastructure equipment. In parallel, a traffic simulation implementation of the Innovation Corridor was also developed for both planning and validation purposes. This study examined how the CV Eco-Approach and Departure (EAD) application would perform by using two instrumented vehicles operating in normal traffic conditions. The fuel consumption and emissions of these vehicles were compared, along with a comparison of two emission models.
The “Innovation Corridor” was set up as a testbed that can be used for CV testing. All of the traffic signal controllers along this corridor were upgraded to be compatible with Society of Automotive Engineers (SAE) connectivity standards. Dedicated Short-Range Communication (DSRC) roadside-units were installed at three key intersections. At these instrumented intersections, Signal Phase and Time (SPaT) messages were directly transmitted from the DSRC roadside units and could be received by vehicles equipped with onboard communication equipment. In addition, Radio Technical Commission for Maritime Services (RTCM) position messages and intersection MAP messages were broadcasted via the DSRC roadside units to support geofencing and accurate vehicle positioning. With this initial instrumentation, CV experimentation was conducted that utilize the SPaT data from these intersections to stabilize traffic flow and reduce emissions.
Prior to the real-world experiments, traffic simulation model runs were carried out on a “virtual” testbed that was developed to imitate the same University Avenue corridor using the high-fidelity microscopic traffic simulation model VISSIM.
For the real-world testing on the Innovation Corridor, two vehicles were tested simultaneously. One test vehicle was utilized that fully implemented the connected vehicle EAD application, while the other vehicle was used as a comparison vehicle, driven normally with traffic without the EAD application. The experiments were conducted during the middle of the day (i.e., between 10:00AM and noon, and 1:30PM to 3:30PM) on a typical weekday. During the experiments, the actual fuel consumption from the vehicles were recorded in real-time, along with detailed trajectory information (i.e., vehicle speed and position at 1 Hz). Once the vehicle trajectories were collected, they were used as input for two different vehicle emission models that were used to estimate fuel consumption and emissions activity: the Motor Vehicles Emission Simulator (MOVES), developed and maintained by the U.S. Environmental Protection Agency, and the Comprehensive Modal Emission Model (CMEM), which was originally developed as an NCHRP project. Both of these emissions models were specifically calibrated for the light-duty test vehicles.
- Results from simulation indicated an 11.46 percent and 30.43 percent reduction in CO2 emissions due to EAD-equipped vehicles, as estimated by MOVES and CMEM, respectively.
- Similarly, simulation results showed an 11.5 percent and 30.4 percent reduction in fuel consumption due to EAD-equipped vehicles, estimated by MOVES and CMEM, respectively.
- The study stated that in the real world, the driver was given the optimized trajectories as advice, and typically had trouble following the speed trajectories exactly. Therefore, it was expected that the simulation results would always be higher than what was measured in the real-world.
- The results from the real-world testbed revealed the following results: 6.6 percent reduction in CO2 emissions, and 6.63 percent reduction in fuel consumption.
- Further, the CMEM estimation methodology for the real-world experiment gave a 4.6 percent improvement in terms of fuel consumption, and the MOVES estimation methodology gave a 2.6 percent improvement. One of the reasons why the MOVES model might be underestimating the overall improvement was due to its binning modeling approach.