MRI: Development of iCAVE2 (Instrument for Connected and Autonomous Vehicle Evaluation and Experimentation)

点击翻译按钮获取中文摘要

This project, developing an integrated 5-in-1 instrument for connected and autonomous vehicle (respectively, CV and AV) evaluation and experimentation (iCAVE2), addresses open research challenges in surface transportation systems where statistics on accidents and fatalities, congestion, fuel consumptions, and emissions have raised serious concerns. This work addresses a critical societal need related to evaluating and experimenting with the emerging and potentially transformative CV/AV technologies by developing a necessary instrument that can bridge the gap between existing simulators and road testing facilities. The instrument is useful to researchers in academia and IT industry, and developers and decision makers in the auto-manufacturing, auto-insurance and government transportation agencies. Graduate assistants will be trained with hands-on experiences and multi-disciplinary knowledge. The instrument enables expansion of several programs and contributes to the education and training in the STEM fields. Consequently, better driver training and related rehabilitation should be expected.iCAVE consists of: 1. Multiple driving simulators (DS) to conduct human/hardware-in-the-loop experiments in a Virtual Reality (VR), 2. A traffic simulator (TS) whose vehicles can mimic, CVs and AVs move in mixed traffic situations with a fairly realistic traffic environment, 3. A Network Simulator (NS) with realistic wireless communication models to evaluate various Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communications technologies, 4. Several Instrumented Vehicles (IVs) including a fully instrumented CV and a AV to conduct real-world experiments and connect data, and 5. An Instrumented Environment (IE) with various traffic sensors and road-side units for V3V/V2I communications covering an area of approximately1.2 square miles with representative urban traffic. The instrumented environment and vehicles provide real-world measurement data to build and calibrate parameters/models for use by the integrated simulators. They can also be used to run CV/AV algorithms and applications to collect data, and evaluate their performance and effectiveness. The instrument will be available to both local and remote users. Its main software module will be developed based on a framework for real-time data distribution, with open APIs and as an open-source project. New designs, technologies, infrastructure, and applications will be evaluated and validated before deployment. This flexible, scalable, safe, and realistic platform, expected to be particularly suitable for answering various "what-if" questions related to safety, efficiency, and sustainability arising from human-automation interactions with not-yet-available technologies and rare/extreme events, is the first-of-its-kind with unprecedented capabilities, not offered by any simulator-based instrument or test-beds in academic, industrial or government-based R&D laboratories. The instrument should also enable activities related to Big Data transportation systems.