Autonomous driving reshapes cities: from technical testing to system reconstruction

In April 2026, test vehicles without steering wheels appeared on public roads in London for the first time. This event marked the official transition of autonomous driving technology from closed testing grounds into one of the most complex urban environments. Almost simultaneously, autonomous electric micro-transit buses began operating on the streets of Los Angeles as part of a shared mobility network. These are not isolated experiments but signals of a structural transformation underway in global urban transportation systems.

Real-Time Decision-Making Systems on Urban Roads

Modern autonomous vehicles are equipped with radar, lidar, and high-definition cameras, forming a continuous 360-degree perception network. The vehicle's onboard computing platform processes gigabytes of data per second, plotting routes in real time, predicting the behavior of pedestrians and other vehicles, and making instantaneous decisions to avoid obstacles or accelerate. This capability is no longer merely an integration of sensors and algorithms; it transforms urban traffic into a distributed real-time control network—each AV serves as an edge node of this network.

The key point is that these systems are evolving from a "perception-decision" closed loop into artificial intelligence capable of understanding the urban contextual environment. For example, in the London tests, vehicles had to navigate narrow streets, irregular intersections, bicycle flows, and temporary roadblocks. This requires the system not only to recognize objects but also to understand local unwritten traffic rules and scene intentions. This is the frontier where machine learning and traditional AI converge: making the decision-making process more transparent and explainable, thereby gaining the trust of regulators and the public.

Collaborative Networks: A Bridge from Lab to City Scale

The large-scale deployment of autonomous driving is not the achievement of a single company. The 2025 pilot of autonomous electric micro-transit buses in Los Angeles is a typical case of tripartite collaboration among automakers, software platforms, and mobility service providers. Traditional automakers contribute manufacturing experience and safety standards, tech companies contribute algorithms and data processing capabilities, and mobility platforms handle user access and route optimization. This cooperation model is becoming the standard: by the end of 2026, MOIA America and Uber plan to deploy autonomous shared vehicles based on the Volkswagen ID.Buzz in Los Angeles, further testing the "autonomous driving as a service" business model.

The significance of these collaborations extends beyond the business level. They are effectively building an operating system for urban transportation—a digital infrastructure composed of vehicles, cloud platforms, roadside units, and traffic management centers. The future city will no longer be a simple "roads + vehicles" concept but a mobile ecosystem with unified data interfaces.

Regulatory Fragmentation and Public Trust: Two Major Constraining Variables

However, technology is not the only determining factor. One of the biggest obstacles to the current promotion of autonomous driving is regulatory fragmentation. Different countries and cities have vastly different regulations on testing permits, safety standards, and liability attribution. For instance, the London tests received special approval from the UK government, while the Los Angeles pilot follows the strict framework of the California Department of Motor Vehicles. This inconsistency forces companies to adapt individually for each market, slowing the scaling process.

Public trust is an even more subtle challenge.Public trust is a more subtle challenge. Surveys indicate that a significant proportion of residents are hesitant about riding in autonomous vehicles. Changing this mindset requires sustained public education and a visible safety record. Fortunately, as more success stories accumulate—including records of zero-accident operations and efficient emergency response demonstrations—public acceptance is slowly increasing. The role of city managers is not only that of regulators, but also of communicators, who need to build platforms for residents to experience and give feedback firsthand, thereby making the technology transparent.

The Hidden Reconstruction of Infrastructure

The impact of autonomous driving extends far beyond the vehicles themselves. Cities are beginning to rethink street design: whether to retain traditional traffic signals or replace them with vehicle-to-road communication. As private car ownership declines, a large number of parking spaces will be freed up, which can be converted into parks, bike lanes, or micro-logistics hubs. Pilot projects in Los Angeles have already combined charging stations with shared docking points, evolving toward multimodal transportation hubs.

In addition, the massive amount of data generated by autonomous vehicles is being used for macro-level traffic management. By analyzing this data, urban planners can precisely optimize traffic light timing, detect road hazards, and predict congestion patterns. This marks a shift from "statistics-based planning" to "real-time data-driven dynamic regulation" in urban governance.

Future Vision: From Taxis to Urban Public Services

The ultimate form of autonomous driving may go beyond personal mobility. Currently tested autonomous delivery vehicles, self-driving emergency vehicles, and dynamically routed shuttle buses suggest that this technology could become a universal platform for urban public services. When autonomous buses can adjust their routes and schedules based on real-time demand, the efficiency and coverage of public transportation will leap forward, especially benefiting currently underserved areas.

In the longer term, autonomous driving will drive cities toward a "de-automobility" evolution. When transportation costs decrease, safety improves, and sharing models mature, cities can gradually reduce dependence on private cars and free up more public space. This is not just a transportation revolution, but an evolution of urban form.

Conclusion

Autonomous driving is becoming part of the city's digital nervous system. The four gears—technology maturity, establishment of cooperation networks, refinement of regulatory frameworks, and building of public trust—must rotate in sync. The practices in London and Los Angeles in 2026 show that we are at a tipping point from "testing" to "system integration." In the next decade, cities will no longer just be places that accommodate autonomous vehicles, but will co-evolve with them to create a new mobility civilization.