AVs Are A National Issue Now
Plus, Tesla Cybercab is actually the most efficient EV ever, and building a self-driving wheelchair for the disabled.
This is the Ride AI newsletter: The most comprehensive weekly digest of news and intelligence at the intersection of technology and transportation.
ICYMI, early bird tickets for Ride AI 2027 are available. If you’d like to apply for an early spot, please do so here.
Top Story
We’re seeing a shift in how AV services present themselves, as Waymo has aired its first-ever national TV commercial, the first time an autonomous-vehicle company has pitched itself to the national public on broadcast television beyond the tech press and early-adopters.
The spot leans entirely on safety, claiming the Waymo Driver is statistically 10x safer than a human driver in the cities it serves. The move says a lot about where Waymo thinks it is. A national ad buy treats robotaxis as a mainstream consumer product, which, with over 10 cities launched already, Waymo increasingly is.
Here at Ride AI, we are of the belief that as AV technology matures and becomes more prevalent, the barrier for adoption becomes marketing, rather than technology. James’s opening remarks at our 2026 conference highlights this problem perfectly. If you haven’t seen it, I think it’s worth a watch.
The national Waymo ad landed the same week Waymo rolled out a paid loyalty program, both signs it’s optimizing for a bigger, more habitual rider base.
Domestic News
The national Waymo ad landed the same week Waymo rolled out a paid priority status program, Waymo Premier, signaling the company’s strategy optimizing for a bigger, more habitual rider base.
Premier is invite-only, costs $29.99 a month, and is only available for Waymo’s most frequent riders (for now). Members get priority pickups (skipping the queue), 10% Waymo Cash back on every trip and more during busy times, early access to Waymo in new cities, and up to five free cancellations a month. It’s launching first in San Francisco, Los Angeles, and Phoenix, with benefits that follow members city to city. Similar to priority programs from airlines and other ride-hailing services, it’s a signal Waymo now has enough repeat riders to bother monetizing loyalty.
Will it get you early access to the newly deployed Ojai vehicles, however? Signs point to no.
Waymo has also published ReD (Reference Driver), a new benchmark for how a careful, competent human avoids a crash, in joint research with TU Delft now out in Nature Communications.
ReD attempts to simulate how a driver updates their beliefs as a situation unfolds, manages uncertainty about what other road users will do, and picks an evasive move (brake, swerve, or both). It builds on Waymo’s earlier NIEON model and treats human driving as the minimization of “surprise.” Unlike models that only capture last-second reactions, ReD also models proactive avoidance.
ReD gives Waymo, and potentially the whole industry, an automated, peer-reviewed yardstick for the question behind every AV crash: “Would a good human driver have avoided it?” Waymo is releasing the research code under an academic license and lobbying SAE and regulators to adopt it.
Tesla applied for a Nevada permit to run up to 5,000 robotaxis across the Las Vegas area in its first year of operation.
The application covers Clark County, Las Vegas, Henderson, and both Harry Reid International and Henderson Executive airports. 5,000 is a huge number. However, Tesla’s actual driverless fleet in other regions is tiny, roughly 20 vehicles in Austin, so 5,000 seems to potentially be a deployment for the upcoming mass-produced Cybercab, pending its validation. Protests to the application are due by July 5, and Tesla requires a full sign-off from Nevada regulators.
Love it or hate it, the Cybercab is the most efficient EV ever produced, per newly surfaced EPA certification documents that lay out its full specs for the first time.
The filing pins the Cybercab at a 165 Wh/mi efficiency rating, with a 219-HP front-wheel-drive motor, a 47.6 kWh pack, and roughly 293 miles of adjusted range. Though, at 3,113 lbs it’s heavier than you’d expect for a two-seater with no steering wheel or pedals (a Mazda Miata weighs about 2,341 lbs), and the front-wheel-drive layout is a first for Tesla. Charging is primarily wireless induction, but efficiency loss has not been officially disclosed or measured. The Cybercab’s certification is a big step towards Tesla having a purpose-built driverless robotaxi fleet, something detractors have long held would never be the case. However, the company still has to be approved for a steering wheel exemption for the vehicle, which NHTSA limits to 2,500 vehicles a year, and which only Zoox has been approved for so far.
Waymo bought Apple’s abandoned self-driving proving ground in Wittmann, Arizona for $220 million.
The 5,500-acre site (a 115-acre fake city, a four-mile oval, a dedicated freeway course, and vehicle-dynamics pads) was the centerpiece of Apple’s dead car project. Apple paid about $125 million for it in 2021, so Waymo paid nearly double. Waymo says it’ll use the track for rider-only testing, motion-control work, and operational training as it pushes toward a goal of 1 million weekly rides by year-end.
Separately, Waymo signed a multi-year fleet-operations partnership with Element Fleet Management, the world’s largest publicly traded fleet manager (over 1.5 million vehicles under management), to run the operational backbone of its robotaxi fleet. Through its Element Mobility arm, Element will handle vehicle lifecycle management, charging and energy, maintenance coordination, and day-to-day fleet optimization to keep cars ready and reliable, while Waymo keeps the rider app and full responsibility for the Waymo Driver’s performance. It starts in San Diego and expands to more markets over time, and it’s an interesting move. Waymo already has a deal with Moov to perform the same fleet functions. Is Waymo running the same playbook that it’s been running with ride-hailing network providers? Is the company systematically testing fleet managers until it finds the right match?
International News
Bolt, Pony.ai, and Stellantis will launch an autonomous-mobility pilot in Luxembourg, testing Pony’s seventh-generation vehicles on the country’s roads with a goal of driverless readiness by the end of the program.
Stellantis supplies a midsize van built on its L4-Ready Platform, Pony provides the autonomy stack, and Bolt brings the ride-hailing marketplace. Bolt pitches it as a public-facing “living lab,” leaning on Luxembourg’s relatively permissive testing rules. Bolt is the largest independent, European-founded ride-hailing platform, with 50+ countries and over 200 million customers.
This is another European push for Pony, which recently launched commercial service in Zagreb with Uber and Verne. It also shows Stellantis hedging across AV partners, since it’s separately tied to Wayve. Now, Europe has multiple platforms providing robotaxis, with Uber, Bolt, Verne, and other regional providers making up a web of marketplaces.
Did Tesla present misleading Full Self-Driving safety data to European regulators? That’s the question raised by correspondence Reuters obtained through public-records requests, which shows Tesla pitched regulators in Sweden and the Netherlands with its own safety statistics while lobbying for approval.
The pitch claimed FSD could have “saved 32,000 lives” (a figure that assumes every vehicle in the US, trucks and motorcycles included, gets replaced by an FSD Tesla) and that Teslas travel more than seven times farther between crashes than the average US driver. Critics call the math apples-to-oranges, because Tesla compares airbag-deployment crashes in relatively new Teslas against all reported crashes in the average US vehicle (about 12 years old). 10 of 11 independent researchers who reviewed the methodology called it marketing, not a serious safety study, and the European Transport Safety Council said that if the data is real, Tesla should submit it for independent peer review.
Tesla, on the other hand, argues FSD demonstrably cuts crashes, and it makes the same case publicly, with its EMEA account claiming FSD Supervised had been “over 3x safer than manual driving on Dutch roads” on June 9. Some regulators have been receptive (a Greek official cited data showing FSD “leads to a very significant drop in accidents”), and crucially the approvals haven’t hinged on Tesla’s numbers. The Dutch authority (RDW) approved FSD in April after its own road and track testing, and says it doesn’t rely on marketing claims or external statistics.
What’s the truth? It seems to me that while Tesla’s framing may oversell the safety case, the country-by-country approvals are riding on regulators’ own evaluations and not only Tesla’s presented data.
Heartwarming DIY Wheelchair Autonomy for the Disabled
A maker studio out of China has built a lidar-based self-driving wheelchair for a disabled viewer. I don’t usually include videos like this in this newsletter, but I believe the framing of this video makes it a worthwhile watch. Autonomous technology has the ability to restore personal autonomy to people with disabilities, and this case is a prime example. Also, the production value is absolutely through the roof. English subtitles are available in the subtitles menu. Let me know what you think!
Alright, that’s it from me… until next week. If you enjoy this newsletter, share it with your friend, colleague, or boss. Thank you for reading; Sophia out!