Opinion: AVs Can Do More Than Just Serve People Who Can Afford A Cab
The autonomous vehicle industry drove onto the scene with resources no transportation industry had ever enjoyed before: billions in capital, the most-sophisticated engineering talent in the world, genuine public excitement, and a regulatory environment that laid down smooth asphalt. For a window of time, the dream of redesigning public transportation from the ground up was genuinely within reach.
But, for the most part, the industry has used it to build a better taxi.
Most public scrutiny around autonomous vehicles has centered on whether the technology works and its various mishaps and misdeeds. Did a Waymo just run a red light? Did Tesla Autopilot cause a crash? Are regulators keeping pace with what’s happening on the roads? This focus misses the larger problem. Technically, the vehicles work well enough, helping to prevent crashes and save lives.
Practically, what has emerged is an industry trend that prioritizes hype instead of mobility equity.
Robotaxis remain operational in narrow geofenced corridors across a handful of major cities, serving riders who already have multiple ways to get around, not to mention Ubers, Lyfts, yellow cabs, etc. Yet 45 percent of the U.S. population has little to no access to adequate public transportation, a figure that has barely moved despite years of industry expansion and billions in cumulative investment. Rather than closing that gap, the AV industry has driven away from it.
The problem runs deeper than simple oversight or neglect. Autonomous vehicles actually exacerbate the problem as robotaxis generate “deadhead” miles at scale, with empty vehicles circling between rides and adding congestion to urban streets without moving a single additional person anywhere. In 2025, deadhead miles accounted for nearly half of Waymo’s total travel in San Francisco, according to California’s Public Utilities Commission. They didn’t contribute new mobility options to the city, only additional traffic competing with transit infrastructure already struggling to function.
Meanwhile, the communities most in need of new mobility options are watching their existing ones disappear. Transit agencies across the country are cutting routes and reducing service hours, not because demand has fallen, but because running low-density corridors, early-morning services, and last-mile connections to transit hubs simply costs too much to justify on current budgets. Routes on low-density corridors are always the first to go when finances tighten, and they are the ones that people with the fewest alternatives depend on most. Nevertheless, the AV industry, flush with capital and engineering capacity, has treated this as someone else’s problem.
Yet, this is precisely where autonomous vehicle economics should change the outcome. The financial case for cutting a transit route rests most heavily on staffing costs. Transportation providers continue to report a persistent bus driver shortage, with one in four transit workers worldwide expected to retire by 2035. Many systems are already operating at a fraction of their required driver capacity, forcing route cuts even where ridership demand exists. At the same time, drivers are expensive, and overnight shifts on low-ridership corridors produce unit economics that no transit agency can defend when facing a budget shortfall. Remove the staffing cost, and the calculus shifts substantially. Without drivers to pay or depots to man in the early hours of the day, a bus running at 5 a.m. on a sparse suburban corridor stops being a financial liability and becomes a service an agency can afford to sustain. Routes that transit operators couldn’t justify keeping become routes they can afford to launch.
The evidence that this works is already accumulating. Driverless shuttles are being deployed along Atlanta’s BeltLine, connecting MARTA rail stations, university campuses, and the Lee and White district on fixed short routes designed specifically to close first-and-last-mile gaps that have long frustrated commuters. In Europe, an EU-backed initiative has launched autonomous transit trials in Oslo and Geneva, focused on integrating demand-responsive driverless vehicles directly into existing public transport networks. What remains unresolved is whether the broader industry will drive down the road where the evidence already leads.
The next phase of AV deployment is being negotiated now, in conversations among technology companies, regulators, and transit authorities, assessing whether this technology has anything practical to offer their networks. Transit operators are resource-constrained and not inclined toward optimism. They need a concrete and near-term return-on-investment case, not a promise of transformation. Years of industry effort have gone into building that case for premium riders in high-density ZIP codes. Building it for the agencies that serve everyone else has barely begun.
Cities that move more people more efficiently generate more economic output and more equitable access to healthcare, education, and employment opportunities. A robotaxi serving upscale passengers in a handful of city blocks will not change those numbers at any meaningful scale. Autonomous vehicle technology is already built for public transit and already operating on public roads. The driver may have left the vehicle, but the industry still has to decide what purpose that vehicle will serve.
Read More:
Streetsblog has migrated to a new comment system. New commenters can register directly in the comments section of any article. Returning commenters: your previous comments and display name have been preserved, but you'll need to reclaim your account by clicking "Forgot your password?" on the sign-in form, entering your email, and following the verification link to set a new password — this is required because passwords could not be carried over during the migration. For questions, contact tips@streetsblog.org.
