Parallel Transportation in TransVerse: From Foundation Models to DeCAST
Chen Zhao,Xiao Wang,Yisheng Lv,Yonglin Tian,Yilun Lin,Fei-Yue Wang
DOI: https://doi.org/10.1109/tits.2023.3311585
IF: 8.5
2023-01-01
IEEE Transactions on Intelligent Transportation Systems
Abstract:Rapid development of AI technologies has propelled the seamless integration of physical and cyber worlds with various kinds of online/offline information collected from millions of multimodal sensing systems. The complexity, diversity and uncertainty inherited in such systems, such as Intelligent Transportation Systems (ITSs), have gone far beyond human capacity of managing and controlling. Our team is among the first to propose the idea of utilizing the nearly unlimited computational resources in cyberspace to construct a bottom-up and top-down combined artificial ITSs for testing, experimenting, representation, verification, and validation of physical ITSs. Especially, the parallel transportation has been developed for safer, smarter, greener, and more reliable transportation services. After three decades of research and field studies, the DeCAST in Transverse, i.e., Decentralized/Distributed Autonomous Operations/Organizations (DAO) in transportation systems, has been envisioned. In this paper, we introduce its architecture, operational processes, software and hardware platforms, and real world applications. Specifically, a transportation foundation model driven by artificial transportation systems, parallel learning and federated intelligence, named TengYun, is outlined for DeCAST.
engineering, electrical & electronic,transportation science & technology, civil