At the same time, the country places great importance on the development of digital twins in industries such as transportation. In July 2019, the Ministry of Transport issued the "Outline of Digital Transportation Development Plan," which pointed out that "a modern transportation system should be built with data as the key element and core driver, promoting the continuous integration and interaction of transportation activities in physical and virtual spaces," providing a good policy environment for the application of digital twins in the transportation field.

In October 2021, the Ministry of Transport issued the "14th Five-Year Plan for Digital Transportation Development," proposing six goals: "digital perception of transportation facilities, widespread coverage of information networks, convenient and intelligent transportation services, online collaboration in industry governance, active innovation in technology applications, and strong security guarantees for network safety." The digital transportation system proposed in the policy document is highly consistent with the concept of "digital twins," indicating that the development of intelligent transportation in China is gradually entering a rapid development era of digital twins.

Digital twin transportation is a current trend and hotspot in the industry, but there are several difficulties and bottlenecks in its development. First, there is no unified definition of the connotation and focus of digital twin transportation, leading to overly broad or narrow boundaries. Second, there is a lack of unified understanding of the functions and application scenarios of digital twin transportation; some researchers believe that digital twins can solve all problems, while others think that digital twins can only achieve visual display and are useless in actual business. Third, there is insufficient emphasis on the activities and spatial movements of people and objects in digital twin transportation, overly simplifying digital twin transportation to the digital modeling of transportation facilities. Based on this, I will briefly introduce the above issues and insights based on my personal research practice. Due to constraints of personal ability and experience, there may inevitably be biases and shortcomings, and I hope to engage in in-depth exchanges with industry colleagues to promote the development of digital twin transportation.

The earliest conceptual model of digital twins was proposed by PLM consultant Dr. Michael Grieves, and was then referred to as the "mirror space model." It was later extended by NASA to mean "creating a mirror of physical products in digital space, reflecting the entire lifecycle process of physical entities through digital means."

In summary, digital twins refer to the use of information technology to reshape the real operating state of the physical world in a computer virtual digital space, including static spatial scenes and the movement of people and objects in space. Digital twins not only present a mirror of the real world but also depict the inherent evolution laws of things, describe the mathematical logical deduction relationships between things, and thus can simulate, deduce, and predict the real world.

The concept of digital twin transportation is based on real transportation, incorporating historical and real-time collected traffic data into the established traffic model simulation system using digital twin technology, enabling rapid data fusion and simulation deduction to construct a complete virtual digital mapping of the transportation system. Through big data, artificial intelligence, and traffic simulation technology, traffic optimization plans are generated, and the pros and cons of future plans are reasonably evaluated in advance.

Digital twin transportation has certain particularities compared to digital twins in other industries. Generally, the application of digital twins in other industries maps the real world to digital space through a data layer and conducts a series of business applications based on this. However, transportation also needs to restore the essence of transportation operations through digital twins. The essence of transportation is the movement patterns of people and objects in both spatial and temporal dimensions. Conventional data perception methods based on IoT monitoring can only collect the state of traffic operations rather than the essence of traffic operations. For example, in the case of road congestion, monitoring methods can only perceive road speed to determine whether congestion exists, but it is difficult to directly feedback the reasons for congestion or how the vehicles on the congested road are routed.

Restoring the entire travel pattern of transportation through digital twin transportation requires reasonable attribution modeling based on existing perception data, followed by simulation deduction to support traffic control and governance. However, IoT sensing devices or floating car data can only perceive and obtain data from certain road sections or travel samples, making it impossible to present a full-scene digital twin. Therefore, based on digital twin technology, collecting and integrating multi-source data and conducting attribution modeling to establish a scientifically reasonable simulation deduction platform is a core issue that digital twin transportation needs to address.

Digital twin transportation has four main technology points: "facility foundation," "data perception," "model simulation," and "business applications."

Facility Foundation. Digital twin transportation models and presents static transportation infrastructure and high-precision map networks within the digital twin platform, including elements such as roads, bridges, tunnels, interchanges, and signal intersections. The digital twin transportation foundation in Shenzhen integrates high-definition remote sensing satellite images of 2000 square kilometers, over 6000 kilometers of urban roads, 419 kilometers of rail operation mileage, 27 major hub terminals, information on 650,000 urban business entities, 100G building road BIM models, and a database of 2.6 million housing "land-building-house-right" systems, covering the entire city's road and rail network and building plot information.

Data Perception. Data perception is based on the facility foundation, mapping real-time traffic operation perception data through a set of high-precision map road networks, and managing massive dynamic and static traffic operation data according to the same network rules. Relying on real-time IoT data for perception monitoring and early warning of infrastructure, this is a digital maintenance application of facilities and a major application feature of digital twin transportation.

Model Simulation. Simulation modeling is based on a unified facility foundation and multi-level perception data, restoring the travel methods, paths, and driving behaviors of all traffic participants in real-time, aiming to uncover the individual travel behavior patterns in real environments.
Source:Saiwen Transportation Network