How much will robotaxis really reshape cities?

Author: Bern Grush

Date Published: September 20, 2025

While hardly the tsunami hyped a decade ago, the driverless vehicle (r)evolution is here. Forget the promise of almost all personal travel in robotaxis by 2030,[1] the real transformation will be more nuanced, uneven, and ultimately more complex for planners than driverless evangelists predicted.

Robotaxi adoption will be driven by many factors and enabled by multiple circumstances. Each person who switches from car ownership to reliance on robotaxis will have their own reasons.

Dominos

The order of disruption

The first wave of serious robotaxi users, unlike the curious tourists posting Waymo joy-rides in San Francisco, are young urban professionals who already rely on rideshare services. They are the car-free city dwellers already tired of expensive, unreliable human-driven alternatives or long weary of undesirable public transit services. (I write from experience: I have been car-free since 2019.)

Traditional taxis and ride-hailing services face extinction in second-wave robotaxi markets within a couple of years of launch in each local market. You can be sure that the push-back by the traditional taxi and ride-hail industry in New York City will be fierce,[2] but the economics and expectations are brutal: no driver wages, 24/7 availability, greater privacy, and consistent service quality. A last few holdout taxi drivers will persist to serve niche customers who refuse to use robotaxis out of fear or on principle. This will eventually become irrelevant with the possible exception of special vehicles for the disability community.

Bus riders are next, but only after robotaxi prices fall considerably and as the availability of public transportation erodes in any particular city. SEPTA's current struggles in Philadelphia offer a haunting preview.[3] Public transportation death spirals will feed fresh business to robotaxi operators who drop prices to feed that spiral.

Current car owners will be the last to convert—they overvalue their cars, see them as a sunk cost, and have sticky-ownership habits. Most will purchase another car or two, and wait until robotaxi services are a fluid, reliable service—Comfortable, Affordable, Fast, and Instant (CAFI)—in all circumstances. They will also be much slower to abandon ownership, waiting until they become empty-nesters, retire or move.

Public transit

Here's where predictions get murkier. Robotaxis will surely capture some bus ridership, but many governments may resist letting essential public transportation services collapse. Instead, some may subsidize multi-passenger robotaxi access for low-income residents—a more efficient solution than maintaining underutilized bus routes in some areas.

Where dropping robotaxi prices begins to siphon off ridership, transit agencies risk irrelevance if they do not adapt. Complete elimination of public transportation? Unlikely. Political realities and equity concerns will keep some form of public mobility alive, however dramatically it must be rethought. Incorporating subsidized, multi-passenger robotaxis as new forms of public transportation is one kind of rethinking that has been bouncing around for the past decade.

S-Curve

The robotaxi transformation will follow a classic S-curve over a generation or more. The first few years will see adoption among users who do not already own personal vehicles and instead rely on taxi and ride-hail. The real disruption happens in the middle years, as robotaxi networks erode public transportation, achieve critical mass, costs decline significantly, and convenience reaches a tipping point that makes using robotaxis as natural as keeping a car parked out front. In the latter years of this transformation, the market will approach saturation and margins will be minimized, and ownership will plateau to a new equilibrium.

This new equilibrium will differ regionally. Critical factors mean many will not abandon car ownership. Families with children under 16 (25% of households in North America and Europe) are often motivated to maintain at least one vehicle—robotaxis can't guarantee child seats or cleanliness standards that parents demand. Tradespeople need vehicles for tools and equipment. Many rural areas lack the density to make robotaxi networks viable. These constraints create a natural floor for car ownership that pure economic thinking often ignores.

I offer the following projections as indicators of what I see over the next two decades. Since prediction is fool’s sport, you should read this for its directionality rather than its numeric predictions.

North America and Australia (35-45% ownership decline): Suburban sprawl limits robotaxi efficiency, but urban cores could see 60%+ drops in car ownership. High parking costs and insurance expenses create strong economic incentives to abandon personal vehicles.

Europe (40-50% ownership decline): Dense cities, excellent public transit, and punitive car ownership costs accelerate adoption. Countries like Netherlands and Denmark could lead in abandoning ownership.

Asia and Middle East (ownership slowdown and lower peak): Rising middle classes that are still buying their first cars will purchase personal vehicles faster than robotaxi optimization can reduce demand. The growth story will likely overwhelm the efficiency story, at least initially.

South America (ownership slowdown and lower peak): Improving infrastructure and urbanization may support higher adoption. Economic constraints remain significant.

Africa (ownership slowdown and lower peak): With the current lowest per capita ownership, vehicle sharing is highest in Africa. It is possible that robotaxi systems may allow a majority of African citizens to leapfrog the automobile ownership phase in a repeat performance of Africa’s uptake of cellular technology over land lines.

Bottom Line

The robotaxi revolution will have a significant evolutionary component. Personal ownership preferences run deeper than technology alone can disrupt. We'll see dramatic changes in dense urban cores, modest shifts in suburbs, and minimal impact in rural areas.

Predictions of “no one owing a car,” were fantasies a decade ago and still are. The real transformation will be meaningful but measured: robotaxis will complement rather than replace our congested urban transportation ecosystems.

Planners’ dilemma

If this is the path robotaxi diffusion takes, some of the projections for changes in urban land use will accrue. For example, there will be a diminished demand for parking, but it will hardly disappear. There will be a growing demand for pickup and drop-off reservation areas, and that will repurpose some of that parking space. Any downtowns that surrender parking space to greater density will attract more residents and workers, all of which creates transportation demand that naturally self-limits density. Also, any such new developments, tend to overshoot the accompanying plans for transportation space, so that planner dreams of perfect downtowns balancing buildings, humans and transportation will remain elusive.

The more difficult problem is we will now have a world in which there is large numbers of a new kind of transportation vehicle while we still have huge numbers of the old. We don’t know how fast or how dramatically this will change, so we are entering a decades-long era of yet more urban configurational changes than witnessed in the recent couple of decades.

Automated vehicles promise to offer remarkable planning opportunities, but uncertainty in their diffusion timelines will make capturing those opportunities arduous, at best. The future of urban mobility planning must incorporate large numbers of privately-owned vehicles and commercial ride service vehicles in both automated and human-driven formats. And all of those will come in increasing varieties of size and spatial needs for pickup and drop-off, parking and charging.

How much of our public transportation infrastructure can we maintain—especially intra-city rail? If we lose bus routes we lose rail-passengers, as well. If we rely only on small public transit vehicles as represented by robotic taxis and shuttles, many cities will become hopelessly congested.

As this next automobility transformation unfolds, planners face an unprecedented challenge: accommodating automated vehicles while enhancing and preserving walkability and active transportation. The collision between robotaxi infrastructure and human-scale mobility is inevitable, but how we manage that collision will determine whether automated vehicles create better cities or just more complex ones.

This transition will test every assumption about urban mobility planning. Success won't be measured by how many cars disappear, but by whether we can weave together old and new transportation modes into cities that actually work for people. Balancing these competing demands will define urban design for the next generation—and we have no roadmap for getting it right.

[1] Garfield, L. (2017) "Only 20% of Americans will own a car in 15 years, new study finds" https://www.businessinsider.com/no-one-will-own-a-car-in-the-future-2017-5

[2] Clark, N. (2025) "Taxi drivers protest Waymo’s self-driving car tests in New York City" https://spectrumlocalnews.com/hi/hawaii/traffic/2025/08/26/taxi-drivers-protest-waymo-s-self-driving-car-tests-in-new-york-city

[3] Jeans-Gail, S. (2025) SEPTA Cuts to Begin This Sunday https://www.railpassengers.org/happening-now/news/blog/septa-cuts-to-begin-this-sunday/