Getting cities ready for driverless vehicles

Driverless mobility will fundamentally change city networks, but how can these locations prepare for an autonomous future. By Dominique Bonte

Driverless mobility will fundamentally change the way both people and goods move around cities. City governments face important challenges in terms of the extent to which roads and other urban infrastructure will need to be optimised and upgraded to accommodate these new forms of mobility.

Special report: Autonomous vehicle infrastructure

All eyes on vehicle technology

Plenty has been written on how in-vehicle technology will enable driverless mobility, but relatively limited discussion has taken place on how road infrastructure and public spaces more generally can facilitate the deployment and safe operation of autonomous vehicles.

The current consensus seems to be that the onus will be entirely on the vehicle sensors and accompanying software to navigate any type of road infrastructure condition. For example, lane markings can be digitally reconstituted when they are only partially visible or lacking altogether by analysing the positioning of actual vehicles through analysis of historical and real-time data. Driverless vehicles will need to be able to perform reliably and safely within the constraints of and without any prerequisites related to existing road infrastructure, with AI-based sensor analytics software compensating for unclear or lacking infrastructure including absent lane markings.

Many city centres and downtown areas will be exclusively reserved for driverless shared mobility sometime in the future to address congestion, pollution, and safety issues, resulting in an environment de facto reserved for autonomous vehicles

Automakers simply do not want to depend on the existence of any predetermined level of road infrastructure quality or condition and the maintenance efforts required to achieve this. More generally, these attitudes are informed by a wider trend towards virtualised road infrastructure with relevant information either stored as part of map data—for example, maximum speed limits—or generated on the fly by advanced real-time AI-based sensor data analytics.

Dedicated road infrastructure for driverless vehicles?

Another key consideration is whether driverless vehicles should be assigned dedicated lanes. Again, consensus points towards the need to support mixed environments whereby vehicles with high levels of automation engage with vehicles equipped with systems supporting low or zero levels of autonomous operation, the onus again being put on the vehicle software to understand, anticipate and act upon human behavioural modes.

However, dedicated trajectories for AVs have been reserved for driverless shuttles in several cities across the world. Assigning fixed routes and dedicated lanes allows safe operation during the early phases of driverless mobility during which testing and validation is important. On the other hand, it is quite clear that full on-demand mobility will need to abandon the “transit” concept of fixed routes and fixed schedules to maximise rider convenience, operational efficiencies, and profitability.

Mercedes-Benz autonomous driving San Jose pilot
AVs and human-driven vehicles will be expected to operate side-by-side in many locations

On the other hand, many city centres and downtown areas will be exclusively reserved for driverless shared mobility sometime in the future to address congestion, pollution, and safety issues, resulting in an environment de facto reserved for autonomous vehicles. It can be envisioned that optimised infrastructure could be put in place in these restricted areas, including applying special reflective paint for lane markings maximising their visibility for vehicle sensors. However, in the meantime the default situation will consist of mixed traffic with human-driven vehicles engaging with driverless vehicles.

One way the public space will be transformed to accommodate driverless shared mobility is the arrangement of dedicated curbside locations to drop off or pick up people and goods as well as stationing two-wheel micromobility vehicles. This is just one example of how the evolving nature of traffic due to the emergence of automated driving will prompt city governments to reorganise urban public space.

The role of roadside technology and connected infrastructure

The thorny question of how important connected and active road infrastructure will be for enabling driverless and electric mobility is centred around the role and need for vehicle-to-everything (V2X), 5G, smart traffic cameras and wireless charging.

  • 5G and V2X – While both V2X and 5G are considered important enhancements in terms of enabling access to low latency, mission-critical information, software updates, and cyber security fixes for guaranteeing safe operation, driverless vehicles will need to perform reliably and safely, at least temporarily, in the absence of 5G cellular coverage and V2X roadside units for the simple reason public cellular coverage cannot be guaranteed. If anything, 5G will allow enabling a smoother and more comfortable rider experience.
  • Wireless EV charging infrastructure – Either stationary wireless charging stations at discrete locations or continuous charging infrastructure embedded in the road surface will be critical for ensuring uninterrupted operation of smart urban mobility and freight as no humans will be available to hook up charging cables. In the meantime, shared driverless EVs can be recharged at dedicated locations outside the city at night that can be reached autonomously.
  • Roadside edge compute systems – Smart traffic cameras and adaptive traffic lights will be important components to optimise automated mobility solutions and systems, even though much of the critical traffic and speed information will increasingly be made available directly to the vehicles through digital technologies. This could ultimately make digital signage redundant.

A future of pedestrianised cities?

When considering driverless urban mobility, it is important to take into account how future cities will look like. Recently unveiled visionary urban concepts invariably favour large levels of pedestrianisation with regular traffic moved underground via tunnels and hyperloops. Examples include Toyota Woven City in Tokyo, The Line concept in Saudi Arabia, and Singapore’s Tengah district. However, driverless shuttles and Toyota e-Palette-like autonomous commercial delivery, retail and healthcare units will be allowed to mix with pedestrians and micromobility users at ground level.

Camera sensors built into driverless vehicles and drones will also be used as mobile or airborne surveillance entities complementing and, in some cases, even replacing traditional fixed surveillance and CCTV cameras

Urban living will change dramatically, becoming greener, more local, more distributed and with digital lifestyles taking hold permanently. Within this context, mobility will be split between high-speed long-distance transit and local mobility dominated by micro-mobility and driverless smart mobility with main services provided through autonomous mobile units. Drones for both people and freight transport will exploit the third dimension.

This revolution will also have an impact on roadside infrastructure. Camera sensors built into driverless vehicles and drones will also be used as mobile or airborne surveillance entities complementing and, in some cases, even replacing traditional fixed surveillance and CCTV cameras.

New urban living will propel driverless vehicles into new roles and use cases far removed from the current thinking steeped in private ownership, mobility-centric attitudes, and commute-driven centralised urbanisation.


About the author: Dominique Bonte, Managing Director and Vice President, leads ABI Research’s end markets research team