Will robo-taxis cause radical disruption - or gridlock?

Currently, three trends are accelerating transformative change in automotive: zero-emission vehicles, autonomous driving and new mobility models. Autonomous, zero-emission robo-taxis embody all three of these trends, offering the potential for convenient, personalized transport as an attractive alternative to owning a car or using public transport. However, understanding the potential impact of robo-taxis is not straightforward.

Vehicles in general currently generate three issues: congestion, pollution and safety. While autonomous, zero-emission vehicles would reduce or minimize pollution and safety issues, they could increase congestion. Robo-taxis are likely to increase the number of taxi journeys due not only to their convenience, but also their empty collecting trips, which would, in theory, more than double the number of journeys altogether.

However, robo-taxis also provide an opportunity to significantly increase traffic capacity by reducing safe stopping distances between cars. As robots can react almost instantly, it is reasonable to reduce the enforced safe distance while still achieving higher security levels.

To explore this theory, we developed an in-depth mathematical model to simulate the capacity impact of autonomous cars on a micro scale, based on a real-life intersection in Frankfurt. The research looked at the traffic capacity in multiple scenarios, based on two factors:

· The percentage of autonomous and human-driven vehicles involved (that is, 100% autonomous versus 50% autonomous and 50% human)

· Adaptation of traffic rules to maximize the capacity for autonomous vehicles (whether no adaptation, little adaptation or radical adaptation).

The overall results are summarized below.

The key finding from the simulation was that with 100% autonomous driving, along with radical adaptation of traffic rules to suit it, road capacity would increase by a factor of 10 in most situations compared with today. If we then add in further safety margins, we arrive at an increased capacity of between five and 10 times.

By contrast, with 100% autonomous driving and no adaptation of the rules, capacity would shrink by around 25% due to today’s rules around safe stopping distances, which are based on human behaviour. Modest adaptation would also increase capacity, but not by as much.

However, in mixed traffic, in which autonomous vehicles and human drivers share the road with current traffic rules, traffic capacity would also decrease. Humans tend to drive above the speed limit and maintain safety gaps that are too short. As autonomous vehicles automatically obey the law, this noticeably slows down traffic flow in our dynamic simulation – essentially, autonomous vehicles are “bullied” by human drivers. In mixed traffic, adapting traffic rules would only provide a slight increase in capacity.

In summary, the only way to effectively address street capacity problems with autonomous vehicles is to switch to 100% autonomous driving and reduce safety distances between cars. Getting rid of congestion and bringing in the convenience of robo-taxis will be primarily a matter of political and regulatory choice. It will require understanding of each city’s mobility profile, existing vehicle density, geography and traffic systems.

Effectively, the price for getting rid of congestion and enjoying the other benefits of robo-taxis would be to render public transport less attractive and forbid human-driven vehicles, at least at peak times or on peak roads. This would impact the ownership rights and behaviour of millions of people. Our calculations show that new robo-taxi models might be cheaper than existing public transport, even if the latter is subsidized. Cities and societies would therefore come to very different conclusions about what choices to make, based on these three key questions:

· If robo-taxi fleets will be clean, nearly error-free and probably traffic jam-free, aren’t they an ideal mobility solution to replace both public transport and regular cars?

· Is a city willing to exploit this potential by embracing the disruption to norms that it would bring?

· If a city is not willing, how will this impact competitiveness in comparison with those that will opt for this type of mobility and accept its consequences?

While the traffic congestion analysis above assumes that autonomous vehicles will ultimately replace conventional vehicles, it is important to understand whether consumers will actually want to travel in robo-taxis. In 2019, Arthur D. Little conducted a worldwide consumer survey in 13 countries around automotive megatrends. The resulting unique data set provides a valuable indication of the likely demand. We found that:

· Consumers without cars would be likely to use autonomous vehicle-based mobility services to replace journeys by both conventional car and public transport. Nearly half - 45% - of people without regular access to cars stated they would use robo-taxis instead of public transport, with 25% stating they would use them instead of conventional taxis.

· For car owners to switch to robo-taxis, they would need to be convinced that robo-taxis would at least match the experience in the key areas of independence, comfort and convenience, which were the top-three reasons consumers gave in our survey for owning a car. According to our survey, 44% of consumers would replace their cars with autonomous cars and robo-taxis.

Thus, once robo-taxis are available, consumer demand will be high, potentially replacing public transport and owned cars, inducing even more traffic. Only with radical regulation towards autonomous driving will this be possible. In contrast, today’s car-sharing remains niche, with nearly 70% of consumers never car-sharing, and only 12% sharing more than 10 times a month. This lack of appetite can be explained by the need for journeys to be plannable and reliable, especially journeys to and from work, which account for half of all car journeys. Robo-taxis could return by themselves from city centres to pick up other commuters during the same rush hour, in the same way public transport does today. However, as well as autonomous driving advances, this will require investment in intelligent-demand fleet management, including dispatching, advanced e-hailing and active demand steering, if it is to move car-sharing into the mainstream.