Recently, the research team led by Professor Liu Zhiyuan from our school published a paper titled "Urban energy and transport impacts of autonomous robotaxi deployment" in Nature Sustainability. Nature Sustainability is a well-known sub-journal of the Nature portfolio and a top international journal in interdisciplinary fields such as environmental science and ecology, with a latest impact factor of 32.1. This study is the first to systematically propose an urban transport–energy coupling analysis framework for autonomous robotaxis (Robotaxi), advancing related research from theoretical models and simulation-based deduction to empirical diagnosis and collaborative optimization based on city-level real-world operational data. Drawing on real-world data from Wuhan and other locations, it examines the interaction between autonomous intelligent systems and existing urban transport systems, reveals their operational characteristics and energy effects, and provides empirical evidence for the sustainable deployment of autonomous mobility.
Urban transport is a complex giant system formed by the interaction of human travel demand, road networks, public transit, operating platforms, and energy infrastructure. As an embodied intelligent agent capable of autonomous perception, decision-making, and action, Robotaxi enters this system, and its impact will be transmitted to the entire urban transport and energy operation system through service area selection, vehicle dispatching, intermodal connections, and charging behavior. The core scientific question that arises is: How do individual vehicle operational decisions aggregate into system-level impacts at the urban scale, and how do human travel demand and existing infrastructure shape the service boundaries of machines?
To address the above issues, the team constructed an analytical framework of "operational diagnosis–demand modeling–collaborative optimization," systematically revealing for the first time the interaction patterns between Robotaxi and the existing transport system as well as its energy effects. The study found that the demand for Robotaxi and human-driven taxis overlaps spatially, but around subway stations and bus stops, the proportion of station-to-station trips for Robotaxi is significantly lower than that for human-driven taxis, suggesting that Robotaxi may both compete in the on-demand mobility market and have room to form complementary services with public transit.
On this basis, the team further revealed the internal mechanisms of Robotaxi operations from both the supply and demand sides. On the demand side, the study identified the constraining mechanism of road network complexity on the spatial distribution of Robotaxi services, significantly improving demand prediction performance; on the supply side, the study quantified the optimization potential of dynamic ride-pooling coordination for fleet size and energy consumption, establishing a quantitative trade-off between service quality and operational efficiency. This indicates that the transport and environmental benefits of deploying Robotaxi stem not only from improvements in individual vehicle capabilities, but also from whether a large number of vehicles can achieve effective coordination around human travel demand.
Robotaxi is a typical scenario of embodied intelligence entering public space. It serves human mobility and, through algorithms, determines when, where, and in what manner vehicles participate in urban operations. The social value of such technology depends not only on the capabilities of the machines themselves, but also on whether machine behavior can be coordinated with public transit, road resources, and energy systems. Using real-world operational analysis as a fulcrum, this study provides a quantifiable case for understanding the interactive evolution of humans, machines, and complex urban systems, and also offers an analytical foundation for Robotaxi to transition from localized pilot programs to efficient, low-carbon, and governable large-scale operations.
Zelin Wang, a doctoral student from the School of Transportation at Southeast University, is the first author of the paper, and Professor Liu Zhiyuan and Professor Jiang Wei from the School of Electrical Engineering are the co-corresponding authors. Southeast University is the first affiliated institution of the paper, and the first author and both co-corresponding authors are all from Southeast University, making this the first paper in Nature Sustainability in which all first and corresponding authors are from Southeast University. This research was supported by projects including the National Natural Science Foundation of China Young Scientists Fund (Category A) and the Young Student Basic Research Project (Doctoral Students).
Paper link: https://doi.org/10.1038/s41893-026-01944-2
