Cooperative Driving Automation (CDA) Research Program: Development, Testing, and Evaluation
|CDA Program research vehicles driving at the Turner-Fairbank Highway Research Center campus in McLean, VA.|
The Federal Highway Administration’s (FHWA) new Cooperative Driving Automation (CDA) Program has put a spotlight on the entire transportation system and how infrastructure can be leveraged to support mobility and safety.
CDA is poised to transform our Nation’s roadways. This technology will improve the efficiency and safety of vehicles across the transportation system. CDA applications will allow interaction between automated vehicles and nonautomated vehicles, other road users, and transportation infrastructure. The CDA Program consists of focus areas that support research, testing and evaluation, engagement, and safety. CARMA is a continuously developed set of tools within the CDA Program that aid and enable the advancement of the focus areas.
“CDA is essential for leveraging all that new and emerging transportation technologies have to offer. Communication with the infrastructure in conjunction with automation has the greatest chance of significantly improving mobility and safety,” says Dr. Lily Elefteriadou, professor at the University of Florida.
In 2013, FHWA launched CARMA with a focus on studying and developing vehicle to everything (V2X) capabilities and algorithms. In 2015, FHWA was one of the first to demonstrate CDA and made CARMA software open source to encourage collaboration with other entities engaged in the research and development of CDA.
The CDA Program contains focus areas that contribute to creating an integrated and intelligent transportation system where automated vehicles work together.
The CDA Program’s research tracks leverage partnerships with other U.S. Department of Transportation agencies—the Intelligent Joint Program Office, Federal Motor Carrier Safety Administration, Federal Transit Administration, and U.S. Maritime Administration—and explores CDA applications relating to traffic, reliability, and freight. The traffic research track investigates solutions to recurring traffic congestion and aims to demonstrate CDA applications in improving road safety and congestion. The reliability research track examines solutions to nonrecurring traffic congestion, such as road weather management, traffic incident management, and work zone management. Lastly, the freight research track explores CDA applications for commercial vehicles, such as buses and trucks.
The CDA Program also conducts testing and evaluation in either a simulation environment, at a physical testing site, or through a hybrid of those two approaches. CDA simulation testing leverages simulation technologies, such as software-in-the-loop—which allows the testing of source code in a simulation environment to test out CDA features before they are ready to be deployed in a real-world testing environment. CDA physical testing includes full-scale and scaled-down vehicles and is conducted at test sites throughout the United States. CDA hybrid testing combines simulation and physical testing.
The CDA Program takes a two-pronged approach to engagement. This approach intends to accelerate CDA adoption and innovation through CDA Collaborative and Support Services. CDA Collaborative engages with stakeholders to facilitate collaboration, the use of CDA technologies, and form a CARMA software user community. CDA Support Services provide help desk assistance to CARMA users and develop training materials for them.
The CARMA Ecosystem consists of open-source products performing core and support functions to advance the CDA Program’s focus areas. The core products are CARMA Platform, CARMA Messenger, CARMA Cloud, and CARMA Streets. These products provide necessary software for conducting CDA research and testing.
The support components are CARMA Analytics, CARMA Simulation, and CARMA 1Tenth. The support components enable data analysis and decrease the cost of development through simulation and scaled-down testing methods.
CDA Program Updates
Over the past few months, the CDA Program has been engaged in activities involving the testing and development of CDA features and applications.
TSMO Use Case Testing
The CDA Program research team successfully completed testing for two transportation systems management and operations (TSMO) use cases at the Turner-Fairbank Highway Research Center in McLean, VA.
The first use case focuses on vehicle scheduling and CDA trajectory optimization at stop-controlled intersections. In the proof-of-concept testing, two CARMA-equipped vehicles approach a stop-controlled intersection (i.e., an intersection with stop signs) at different times. As the vehicles enter the radius of the intersection roadside unit (RSU), CARMA Streets receives their status and intent and calculates the vehicles’ optimal schedules. Schedules refer to the desirable time at which each vehicle should arrive at the intersection. Using this schedule information, each vehicle adjusts its trajectory and travels through the intersection accordingly.
Transportation systems management and operations testing scenarios
The validation testing tested two scenarios, conflicting and nonconflicting routes. In the conflicting scenario, the two vehicles were routed on paths that would require them to communicate with the RSU and adjust their trajectories to avoid collision and to depart the intersection based on the first-in-first-out rule.
In the nonconflicting scenario, the two vehicles were routed on nonconflicting paths. In this scenario, the starting locations of the vehicles were determined so that they both arrive at the intersection at about the same time. Since their paths had no vehicle-to-vehicle conflicts, CARMA Streets let both vehicles depart from the intersection at the same time, which would improve mobility compared to a non-CDA environment. The second use case involves a signalized intersection with fixed-time traffic signal settings (i.e., intervals for each traffic signal indication are fixed). Two vehicles approach the intersection and receive signal phase and timing (SPaT) messages from an RSU. Each vehicle adjusts its trajectory according to the received SPaT messages and travels through the intersection.
The TSMO use case testing has demonstrated successful validation of CDA features in a closed test track. These use cases aim to ultimately reduce traffic congestion, improve energy efficiency, and increase infrastructure efficiency.
Port Drayage Testing
In December 2021, in addition to TSMO use case testing, the CDA Program team traveled to a closed test track at SunTrax in Auburndale, FL, for port drayage proof-of-concept testing under the freight research track. Port drayage is the short haul of shipping containers between a port and a warehouse; during testing, a staging area served in lieu of an actual warehouse. The team equipped a commercial motor vehicle (CMV) with automation technologies, including CARMA, to demonstrate CDA applications to port infrastructure.
The testing scenario proceeded as follows: a CMV was stationed outside the entrance of the staging area. The automated driving system of the CMV was engaged, and the CMV moved toward the staging area entrance.
|Commercial motor vehicle stationed outside the staging area during a test.|
The CARMA Platform communicated with an RSU interfaced with CARMA Streets, representing the infrastructure at the staging area, to get the container pickup location in the staging area. The CMV determined a route to the container pickup location and moved toward that location. When the CMV arrived at the pickup location, it communicated with the RSU which then communicated with the container handling equipment operator to load the container.
After the container loading was completed (manually), the CMV received a new route to the staging area exit from the RSU. From the staging area exit, the CMV received another route to the port entrance from the RSU within the staging area to move toward the port terminal. Once the CMV arrived at the port entrance, it communicated with an RSU interfaced with CARMA Streets within the mock port terminal and received directions to head to a location to unload the container.
Afterwards, the CMV proceeded to a new pickup location within the mock port area to load another container. The CMV then moved toward an inspection checkpoint after receiving directions from the port infrastructure. If the CMV passed the inspection, it was directed toward the port exit. If the CMV did not pass, it was moved to the holding area where an inspector was waiting to perform an additional inspection. If the CMV passed the additional inspection, the inspector communicated to the port infrastructure that the CMV had passed. From there, the CMV was instructed to exit the port.
The port drayage proof-of-concept testing allowed researchers to demonstrate the capabilities of an automated CMV for port drayage operations and highlighted the benefits of CDA to port management. This proof-of-concept testing also points out the level of reliability wireless communications could provide to facilitate the movement of CMVs performing the complex port drayage process.
Port Drayage Testing.
The products under the CARMA Ecosystem support the cutting-edge research and development performed by the CDA Program. The core component, CARMA Platform, and the support component, CARMA Everything-in-the-Loop (XiL), were recently released.
CARMA Platform enables CDA features to allow automated driving systems (ADS) to interact and cooperate with infrastructure and other vehicles. Version 4.0.3 was released in May 2021 and included feature enhancements and bug fixes. Version 4.0.3 is the first version of CARMA Platform to begin the transition to Robot Operating System 2. Feature enhancements were added to support proof-of-concept applications for TSMO use cases, including cooperative traffic signaling and CMVs in work zones.
The CARMA XiL project has developed a cosimulation tool to support the development, evaluation, and deployment of CDA technology. This tool supports the integration of CARMA Platform cooperative-ADS (C-ADS) features with open-source Cars Learning to Act (CARLA), Simulation of Urban Mobility (SUMO), and NS-3, using the MOSAIC simulator framework to simulate traffic and realistic C-ADS in one package. This cosimulation tool also uses NS-3, a network simulator, to simulate the V2X communications used by the C-ADS systems. The MOSAIC simulator framework manages the interaction between all of these. The CARMA XiL cosimulation tool is leveraged by different CDA use cases research such as traffic incident management and traffic signal priority.
Version 1.0 was released in the third quarter of 2022, and includes the following features:
- CARLA-SUMO integration with the MOSAIC framework.
- CARMA Platform integration with CARLA for a single connected and automated vehicle.
- Vehicle-to-vehicle communication research support.
The following features of cosimulation tool are under development:
- CARMA Platform integration with CARLA for multiple vehicles.
- CARMA Streets and cosimulation tool interface development.
- Vehicle-to-vehicle and vehicle-to-infrastructure communication research support.
Learn more about CARMA XiL and its components:
Over the next few months, the CDA Program will continue various activities to test CDA research and applications. The cooperative perception TSMO use case has begun initial testing at TFHRC. Cooperative perception (CP) allows entities to share data perceived locally, with the aim of improving safety for pedestrians and road users. The CP use case will lay the foundation for future use cases performed under CDA research projects. Additionally, the Integrated Highway Prototype II project will perform testing as well, with the purpose of advancing CDA freeway features.
As the CDA Program continues to develop and test CDA tools and applications, it will accelerate industry adoption of the technology and improve transportation system mobility, safety, and efficiency.
“The FHWA Saxton Transportation Operations Laboratory is conducting critical research that is required to demonstrate the potential of cooperative driving automation to achieve safety, mobility, and environmental impacts,” says Dr. Larry Head, professor at the University of Arizona. “They are pushing the boundaries of the technology and developing new knowledge that will establish the roadmap for the future of automated transportation systems.”
Govind Vadakpat is the CDA Program Manager in FHWA’s Office of Safety and Operations Research and Development, leading the open-source development and collaboration efforts of CARMA with partners and stakeholders. He holds a doctorate in civil engineering from Pennsylvania State University and a master’s degree in civil engineering from the University of Wisconsin–Madison.
Danielle Chou is program manager for emerging technologies in FHWA’s Office of Safety and Operations Research and Development, including the VOICES project. She earned a B.S. and M.S. in mechanical engineering from the Massachusetts Institute of Technology.
Hyungjun Park is a technical program manager of the CDA Freight program in FHWA’s Office of Safety and Operations Research and Development. He manages various CDA activities, focusing on commercial motor vehicles and freight. He earned a B.S. in city planning from Hanyang University in South Korea, and an M.S. and a Ph.D. in civil engineering from the University of Virginia.
Steven Vu is a contracted communications specialist in FHWA’s Saxton Transportation Operations Laboratory, contributing to marketing and outreach activities. He earned his B.A. degree from the University of Virginia.
For more information on the CDA Program, visit https://highways.dot.gov/research/operations/CARMA.