How To Prepare Our Cities for the Autonomous Vehicle Revolution

A few months ago I had the opportunity to take a planned ride in an autonomous vehicle. (Watch my video recap here.) I was in Scottsdale, a suburb just outside Phoenix, Arizona, for a work conference. I needed to get from the airport to the conference resort. There weren’t great public transportation options, so I knew I would need to drive. I didn’t need to go anywhere else that trip, so I only needed a ride there and back. This was the perfect opportunity to test out the Waymo autonomous vehicle ride hailing option for the very first time.

When the white crossover pulled smoothly up to the curb, there was no awkward exchange of greetings, no radio static, and no other humans in the car with me. Just me and the technology of the very near future.

I climbed into the back seat and the vehicle merged into traffic. Within five minutes of the ride, the most profound thing about the experience occurred: it became completely ordinary. I stopped tracking the rotations of the steering wheel and started looking out the window at the architecture of the city itself.

That is when the true reality of the driverless shift clicked into place. For years, public discourse has treated autonomous vehicles (AVs) as an exotic, slow-rolling science experiment. We get bogged down in technical debates over lidar arrays, edge cases, and machine learning models. But as I watched the desert city glide past, I realized we are missing the real story. AVs are fundamentally not a software or automotive milestone. They are more than just the next iteration of the horseless carriage. We didn’t create a better way to get from A to Z. Autonomous vehicles are completely changing the way we interact with the physical world around us.

The proliferation of AVs is about to completely dismantle and reconstitute everything we know about our cities, campuses, parking policies, street architecture, economic equity, insurance reform, accessibility and mobility concerns, and urban form.

The driverless future is no longer a distant horizon line. It is active, it is scaling, and it is hitting public infrastructure at a velocity that will catch cities, campuses, politicians, and government officials completely flat-footed unless they start changing their behavior right now.

Act I: Why Your Next Car Won’t Be a Car at All

For the past ten years, municipal planners have comfortably pushed AV integration into long-range planning buckets. But a look at real-world deployment reveals that the transition from experimental prototype to scaled commercial utility is happening right now. The scaling of rider-only services across multiple cities has moved well beyond localized pilots into a viable, heavy-duty commercial mobility framework.

Consider the sheer scale of current operations. Waymo has transitioned from an engineering curiosity into an infrastructure asset, logging more than 200 million fully driverless miles on public roads and clearing more than 400,000 paid passenger trips every single week. Data from the California Public Utilities Commission demonstrates that consumers are embracing driverless transit at an exponential rate. In December 2025 alone, Californians completed more than 1.2 million rides in commercial self-driving cars, representing a staggering 500% surge across a brief 19-month window.

Platforms are shifting away from standalone vehicle manufacturing to serve as the scalable network layer for autonomy stacks. Uber is actively scaling out its hybrid marketplace, expecting to feature operational robotaxis across 15 cities. To secure this distribution strategy, Uber locked in an agreement to deploy 10,000 fully autonomous Rivian R2 robotaxis in its first phase, with options to purchase up to 40,000 more by 2030. This multi-platform approach is mirrored by an integration with Lucid Motors and Nuro, which will deliver at least 20,000 premium Level 4 SUVs into Uber’s ride-share marketplace over the next six years.

This industrial acceleration is accompanied by a major pivot in how vehicle software and hardware are sourced. The initial expectation was that full vehicle autonomy would arrive through consumer car ownership. However, data from global automotive experts indicates a deep structural divide between private and commercial use cases.

A recent biannual survey conducted by the McKinsey & Company Center for Future Mobility revealed that 49% of industry decision-makers predict that the mass consumer automotive market will anchor on Level 2+ advanced driver-assistance features rather than full vehicle ownership by 2035. This marks a definitive shift away from previous expectations centered on consumer Level 3 systems. The primary barrier is financial. Engineering a Level 4 tech stack that can successfully navigate complex urban edge cases and unstructured environments demands more than $3 billion in software research and validation costs per stack. Per-unit manufacturing costs for these highly specialized vehicles still exceed $100,000. Consequently, automotive original equipment manufacturers (OEMs) are increasingly decoupling software stacks from hardware assembly, choosing a mix-and-match sourcing strategy. This economic reality guarantees that fully autonomous systems will scale first inside commercial, corporate, and municipal fleets where high vehicle utilization can amortize massive development costs.

For years, marketing campaigns hinted that you would soon buy a vehicle, climb into the back seat with a book, and let it drive you to work. These data points show that that future is dead. Because validating fully autonomous software costs billions of dollars and individual vehicles cost over $100,000 to manufacture, automakers cannot afford to sell full autonomy directly to consumers. Your next personal car will not be a sci-fi pod. It will simply have highly advanced driver-assistance features like tighter lane-centering and smarter cruise control. You will still be legally and practically responsible for sitting in the driver's seat with your eyes on the road.

Because full autonomy (Level 4) is scaling exclusively inside commercial, corporate, and city fleets, you will experience true driverless technology as a utility rather than an item you own.  Instead of saving up to buy a self-driving vehicle, you will hail a driverless ride through an app, step onto an autonomous campus shuttle, or watch an automated pod deliver your groceries. The driverless transition will hit your life through the businesses you use and the municipal transit systems your taxes fund, fundamentally shifting your relationship with transportation from property ownership to on-demand access.

Maintaining a private automobile is an immense financial burden, frequently eating up a massive portion of a household budget. The industrial math means that hyper-efficient corporate fleets will soon be able to distribute their massive development costs across millions of rides a day. As the per-mile operating costs for these fleets plunge, riding in a driverless vehicle will become drastically cheaper than owning, insuring, fueling, and parking a private car. For the average person, this financial tipping point opens up a realistic path to an entirely car-free lifestyle without sacrificing personal mobility.

Act II: Macro Safety for All. Micro Safety for Most

The absolute bedrock of the public policy argument for AV integration is a moral imperative centered on public safety. Human drivers are a public health catastrophe. On American roadways alone, human error, distraction, and impairment claim approximately 39,000 to 40,000 lives every single year.

The empirical data logged across the first 127 million driverless miles offers clear evidence of a transportation landscape that removes human frailty from the wheel. Independent safety projects, including the Connected Environment and Smart Intersections Project at the University of Michigan, show a 10-fold reduction in collisions involving serious injuries and a 12-fold drop in pedestrian-involved incidents compared to human benchmarks. Aggregate safety metrics report 81% fewer airbag deployments, 85% fewer severe injury crashes, and 92% fewer pedestrian injury claims across automated mileage pools.

For safety offices and state planners, the baseline of political justification will inevitably pivot from permitting technological innovation to actively executing a non-negotiable public health mandate.

However, an authentic perspective from the decision table requires us to confront a secondary, highly turbulent reality. Macro safety statistics do not automatically prevent micro-operational friction on municipal streets. While robotaxis are statistically superior at obeying literal traffic signals, they are experiencing significant friction when confronting the messy, unmapped realities of active urban management.

This tactical disconnect was demonstrated during a mass-casualty emergency event in downtown City of Austin, Texas. Police Chief Lisa Davis confirmed that officers and paramedics responded to the active scene within 57 seconds. Yet, municipal operations encountered structural blind spots as multiple autonomous vehicles completely disregarded hand-signals from on-ground traffic officers directing active staging zones. First responders highlighted critical instances of self-driving cars driving through flooded roads, disregarding train barricades, and choking transit arteries during active emergency operations. This micro-friction is exacerbated by system-level failures during local network blackouts. A high-profile outage in San Francisco caused scores of autonomous vehicles to stall simultaneously, creating traffic jams that paralyzed localized corridors.

These structural issues have triggered active regulatory investigations by the National Highway Traffic Safety Administration NHTSA into autonomous vehicle software stacks colliding with stationary objects .

In response, local lawmakers are creating direct legislative corrections. California's Assembly Bill 1777 requires all commercial autonomous vehicle platforms to feature real-time, two-way communication override loops connected directly to remote command centers. Public trust moves at the speed of video, and the industry is learning that long-term commercial survival requires solving these complex street-level edge cases.

Act III: The Land-Use Wrecking Ball

While technology commentators focus on computing architectures, urban leaders must understand that vehicle autonomy is fundamentally a land-use narrative. Today’s cities are essentially storage networks for underutilized metal on wheels. In typical major metropolitan cores, single-family zoning paradigms and mandatory parking minimums reserve up to 75% of residential land area for car dependence, suppressing density and inflating structural housing costs.

Autonomous fleets challenge the core spatial math of vehicle storage. Because a commercial robotaxi operates on a high-utilization loop, the structural requirement for static, long-term parking begins to evaporate. McKinsey explicitly notes that fleet operational costs are projected to plummet from over $8 per mile down to just over $1 per mile by 2040 due to vehicle longevity and optimized fleet maintenance software. This hyper-efficient per-mile math gives developers and municipalities an unprecedented opportunity to execute adaptive land-use strategies.

Our current zoning ordinances are often reflections of separations. Separation of land use, or of business types, or worse, of people. Of course, there is another separation used for zoning: between people and cars. Current zoning conditions are the legacy of car storage. In the United States, we routinely build between two and eight parking spaces for every single automobile on the road. When shared autonomous vehicle fleet utilization scales and operating costs shrink, the absolute necessity to park cars at their final destinations dissolves.

Surface parking lots represent the low-hanging fruit of the post-car transition. In real estate development terms, these parcels are prime dirt. Because they lack the complex structural barriers of vertical concrete structures, they can be redeveloped quickly. (Quickly is a relative. Municipal leaders must eliminate minimum parking mandates first.)

Instead of luxury high-rises that require massive capital financing, these flat asphalt fields lend themselves perfectly to medium-density multi-family housing typologies. Think row houses, courtyard apartments, and three-to-four-story cross-laminated timber infill blocks.

Proactive cities are bypassing traditional zoning delays through as-of-right residential conversion policies. For example, the Los Angeles Citywide Adaptive Reuse Ordinance demonstrates how removing minimum unit-size restrictions and locking in existing parking counts can compress development timelines and stimulate project financing.

Repurposing multi-story concrete parking garages is one of the thorniest challenges a city practitioner will face when it comes to rethinking the way we design and build cities because of AVs. These structures are fundamentally engineered like highway bridges rather than occupied buildings. They suffer from low floor-to-floor clearances, structural columns packed tight to optimize stall layouts, sloped driving ramps, and immense floor plates that prevent natural daylight from penetrating the interior core. Fully designing a new parking structure to support eventual human occupancy introduces an upfront structural premium of up to 40%.

For any designer, developer, or jurisdiction currently planning to build a new parking structure, plans should be considered for how to repurpose the structure once the cars are gone. In Tampa, Florida, designers Kohn Pedersen Fox KPF provided the designs for a downtown parking garage structure that could be repurposed by taking out the speed ramps and reusing the flat parking levels for various future needs.

Instead of forcing these dark, heavy concrete shells to become apartments, the most practical approach involves selective demolition and structural gutting. By removing the interior center bays where the driving ramps sit, developers can open up massive lightwells and structural atriums. Or, these structures can be re-engineered into terraced vertical parks, multi-level community hubs, and urban green spaces. The structural frame provides an ideal skeleton for vertical gardens, elevated running tracks, and stormwater retention infrastructure. Converting standalone parking structures into pocket parks breaks up the urban heat island effect, absorbs airborne particulates, and delivers essential public health assets to dense urban cores.

The post-World War II office park complexes are surrounded by vast, underutilized oceans of blacktop, explicitly designed to accommodate peak mid-day commuter demand. As automated networks absorb the traditional nine-to-five commute pattern, these sprawling suburban parking lagoons become obsolete. Suburban municipalities must execute sub-district master planning that treats the entire office park as a single master-planned ecosystem. Rather than executing piecemeal building permits, zoning must be updated to facilitate high-density, mixed-use neighborhoods.

This transition shifts the suburb from a destination you must drive to into a neighborhood you can walk through. It allows suburban communities to capture the energetic, walkable lifestyle historically reserved for historic city centers while utilizing automated fleet networks to solve the foundational regional connectivity gaps.

We are already witnessing the market validation of this transition. Progressive developments, such as Culdesac Tempe, are purposefully designed with zero resident parking infrastructure. The master plan clusters automated vehicle access, fleet storage, and localized EV charging hubs strictly at the exterior edges of the development. The internal parking acres are then reclaimed and built out into walkable, human-scaled spaces featuring interior pedestrian paths, deep courtyards, micro-retail bays, and neighborhood parks where vehicles are physically barred from entering. By substituting vehicle storage with pedestrian courtyards, micromobility hubs, and shared transit access, these communities demonstrate how cities can replace asphalt dead zones with productive, human-centric architecture.

Act IV: Way Mo’ Cars

Without aggressive intervention, AVs may worsen traffic congestion at first, not solve it. This is one of the many paradoxes of autonomous vehicles. The more AV options there are for individuals, the less likely they are to take any other form of transportation. AVs will soon offer the most convenient and least expensive transportation option for the individual user. Moreover, for AVs to work at their fullest potential, every vehicle needs to be an AV. Every. Single. One. That means there cannot be a single vehicle driven by a human. That would require a meteoric shift in how America operates.

In shareholder communications, ride-hailing networks like Uber openly admit that adding autonomous supply to an urban market does not trigger a clean, one-for-one replacement of human-driven cars. Instead, the introduction of highly reliable, driverless vehicles drives massive, compounding category growth across the entire app-taxi sector.

Internal Uber documentation explicitly states that "our network benefits from every incremental unit of supply added in a city. As supply increases, customers find more value because rides become more affordable with faster ETAs. This fact alone gives us considerable conviction that AVs (as a new form of supply) will expand—not shrink—our total addressable market."

In multi-platform launch markets like Austin and Atlanta, overall trip volumes grew significantly across both automated and human-driven choices on the network. In San Francisco, alternative supply additions expanded vehicle miles traveled (VMT) across the entire system, pulling users into cars and away from high-capacity mass transit lanes. This dynamics forces city groups like the San Francisco Bike Coalition, led by Christopher White, to actively fight the reopening of historic car-free avenues like Market Street to autonomous fleets and black car services. Furthermore, fleets generate substantial zero-occupant transit loops, as vehicles cruise empty through central business districts to avoid parking structures, creating a compounding congestion loop for urban networks.

Act V: Equity, Labor, and Direct Policy Intervention

The stakes of the regulatory battle are extraordinarily high for working-class and historically marginalized communities. In modern American car-centric architecture, vehicle ownership functions as an aggressive, regressive tax on social mobility. Statistics confirm that the lowest-earning fifth of American households spend a devastating 38% of their entire net income simply to maintain an operational car. In low-density transit deserts, a single blown head gasket or an inability to pay a skyrocketing insurance premium routinely costs a worker their job and slips their household into structural poverty.

If commercial fleet networks can drop per-mile transit economics down to $1, a lower-income resident can eliminate the financial liability of private car ownership while retaining access to regional job centers. But this baseline of equity will only manifest if public leadership intervenes to prevent companies from cherry-picking wealthy corridors and ignoring less affluent areas.

Because state legislatures and federal safety bodies are moving at a slow pace, local municipal leaders are building their own policy guardrails. A premier case study is unfolding via progressive legislation championed by Washington D.C. Councilmember Charles Allen. Rather than waiting for delayed departmental infrastructure studies, the D.C. framework establishes a highly proactive playbook for active municipal management. The legislation introduces a dedicated Vehicle Miles Traveled (VMT) fee directly applied to commercial robotaxi fleets to tax empty, zero-passenger cruising loops and mitigate congestion. Recognizing the severe labor displacement threats facing traditional gig and taxi operators, revenues generated from these VMT fees are legally funneled into transitioning accounts split between mass transit and a dedicated fund for worker retraining and direct compensation. Moreover, to counter the historical pattern where ride-hailing wait times and coverage are significantly worse in low-income neighborhoods, the framework mandates that AV operators submit binding operational plans that guarantee equal wait times and geometric service distribution. Finally, the policy explicitly ties robotaxi networks back to mass transit backbones, offering calculated discount incentives for rides that plug explicitly into local rail or high-capacity bus networks.

As previously noted, lower income families currently bear a disproportionate financial burden for transportation costs. If jurisdictions choose to replace parking lots in transit-accessible neighborhoods with medium-density housing, they can increase housing supply exactly where essential workers need it most, effectively leveraging technological abundance to anchor structural poverty relief.

Beyond the parked, stationary vehicle issue that cities and campuses will no longer have to contend with, there is the issue of high-volume traffic corridors. These highways, boulevards, and parkways routinely inflict severe respiratory damage on urban neighborhoods, yielding high adult and childhood asthma rates in auto-dominated downtown blocks. Often, these high-trafficked roadways are in low-income areas. By reducing or eliminating the need for these arteries, we can also increase the level of health and safety for those who historically have not seen care and attention.

Act VI: Jurisdictional Fragmentation

The final roadblock to scaling this transition is the systemic crisis of regional regulatory fragmentation. In core economic centers, transportation infrastructure spans a complex web of overlapping county, state, and district lines. As business and regional policy groups like the Washington Board of Trade have pointed out, treating autonomous grid rules as isolated city-by-city pilots introduces a severe competitive disadvantage. A vehicle that navigates under one jurisdiction’s insurance laws cannot be forced to drop its passenger off at a state border due to conflicting local standards.

While jurisdictions like Maryland attempt to push statewide frameworks via pending legislation, neighboring areas like Virginia have delayed commercial autonomous vehicle bills, shifting crucial timelines to future legislative sessions. For an autonomous system to operate with true regional utility, state and district leaders must establish cooperative, cross-border permitting agreements and shared data transparency protocols.

Summary

The next time I’m traveling to a city that offers AVs, I will absolutely take one. It has absolutely replaced rental cars, ride hailing taxis, and even me asking a friend for a lift. (It might even replace public transportation options depending on the city.) The future is here and I’m ready to get on board.

The streetscape is up for structural negotiation for the first time in a century, and the window for reactive governance has closed. Practitioners and regional leaders must move past isolated, city-by-city pilots to establish cooperative, cross-border permitting agreements and shared data transparency protocols. We have a fleeting opportunity to capture an immense real estate surplus, transforming multi-story concrete garages and underutilized suburban office parking lagoons into medium-density affordable housing and restorative public parks. Technology remains fundamentally neutral, but our zoning ordinances and pricing mechanisms are not. City leaders who choose to regulate boldly today will actively reclaim their communities for human vitality, while those who hesitate will simply be left to manage an automated state of permanent gridlock.

Autonomous vehicles will not feel revolutionary when you step inside one. But, if cities and campuses get the right planning steps implemented now, we have the opportunity to revolutionize the way our cities are shaped for the better. The choices we make through targeted policy in the near future will determine whether we leverage this transition to reclaim our urban spaces, or simply automate our traffic jams.