Affordable Cooperative Driving Automation Research and Training

The CARMA software suite is becoming a powerful tool to assist researchers in advancing Cooperative Driving Automation (CDA) strategies for new Transportation Systems Management and Operations (TSMO) use cases. The CARMA software suite is being developed to be open source using agile software development practices and is hosted on GitHub. It is being used by multiple partners both within the DOT (e.g. FMCSA, MARAD) and external to DOT (e.g. Universities of Cincinnati and Utah) to support light and heavy vehicle testing. Since the CARMA software suite is open source, it is encouraging research collaboration with the goal of understanding the benefits of CDA by partnering with industry, academia, and the public sector. However, one particular barrier to entry that turns away organizations desiring to perform CDA research is the high cost of vehicles, sensors, and testing. There are not many entities that can afford the high upfront costs, and even these entities are most often very limited in the type of research under the umbrella of CDA that they can pursue. With the goal of eliminating this barrier to entry, this 3 year project will seek to develop a low-cost approach to CDA research through the use of scaled down robotic cars, roughly one tenth the size of real cars. CARMA 1tenth (C1T) cars will, as much as possible, host the same CARMA software as their larger counterparts while coming at a fraction of the cost (roughly 100x cheaper). After developing C1T cars and trucks by equipping them with the hardware (e.g. mini Lidar and camera) and software (i.e. CARMA Platform) to enable them to participate in CDA, this project will have two primary focus areas: (1) Affordable accelerated CDA feature development and testing and (2) next generation training through a C1T course. The purpose of both focus areas is to accelerate CDA technology, both in the short term and long term. As part of the CDA feature development and testing focus area, the project will place emphasis on testing and developing more features, with more vehicles, more often. The initial emphasis will be on proving out the current existing features with more vehicles (e.g. platooning). After that, emphasis will shift on new features that are harder to develop and test with real vehicles either due to budgetary constraints (i.e. cost of vehicles and testing) or safety constraints. For example, a scenario with multiple features that can be tested is composed of vehicles merging into and out of a large, tightly packed, platoon in the midst of heavy traffic (i.e. in the midst of many other C1T vehicles that are not part of the platoon). This first focus area will also include the following: building cooperative infrastructure (e.g. smart signals, RSUs, etc.), developing test procedures for priority use cases, building a data acquisition system and data processing tools to aid with analysis, and managing C1T open source code on Github. Because of the size of the vehicles, and the reduced need for safety precautions, this project will enable developers to test algorithm and software changes at a moment’s notice. This capability will enable drastic improvement in the CARMA program’s development speed, as well its ability to improve and fine-tune algorithms. As part of the next generation training focus area, the project will develop a course and materials geared towards university students and researchers to enable them to develop their own CDA features using C1T vehicles. The course will provide classes on using the CARMA software suite, as well as developing, testing, and evaluating applications. Outreach to universities, providing webinars, onsite visits and demos, and business development will also be a part of this focus area. As a result of this 3 year effort, (1) the CDA community will grow, (2) CDA development will accelerate, and (3) the testing of CDA features will become more widespread.

• Supplemental Notes:
  • USDOT Research Hub DisplayID 157396

Language

• English

Project

• Status: Active
• Funding: $350000.00
• Sponsor Organizations:

  Intelligent Transportation Systems - Joint Program Office

  ,    
• Managing Organizations:

  Federal Highway Administration

  Mclean, VA  United States  22101
• Project Managers:

  Bujanovic, Pavle

• Start Date: 20210601
• Expected Completion Date: 20230301
• Actual Completion Date: 0
• USDOT Program: Accelerating Deployment
• Subprogram: Program Support

Subject/Index Terms

• TRT Terms: Autonomous vehicle guidance; Autonomous vehicles; Connected vehicles; Cost effectiveness; Intelligent transportation systems; Mobile communication systems; Software; Training
• Subject Areas: Data and Information Technology; Education and Training; Highways; Operations and Traffic Management; Vehicles and Equipment;

Filing Info

• Accession Number: 01863633
• Record Type: Research project
• Source Agency: Department of Transportation
• Files: RIP, USDOT
• Created Date: Nov 15 2022 4:17PM