AV Market Update CW 18
Tesla's robotaxi launch, Waymo's 250K weekly rides, VW Uber partnership, Pony.ai's 70% cost reduction, and US regulatory shifts in autonomous vehicles.
👋 Hello!
This week has been particularly eventful in the autonomous driving world, with major announcements across both US and Chinese markets signaling a dramatic acceleration toward commercialization.
Let's explore the significant developments that shaped the autonomous vehicle industry this past week.
Today's newsletter has 2,700 words and takes about 10 minutes to read.
🚗 Global Race to Commercialize Autonomous Vehicles Intensifies
This week witnessed a seismic shift in the autonomous vehicle landscape as major players across the globe made significant moves that signal the industry's accelerating transition from experimental technology to commercial reality.
🇺🇸 US Market: Competition Heats Up Amid Favorable Regulatory Changes
The US market is rapidly evolving into a full-fledged autonomous mobility battleground with Tesla, Waymo, and a new Volkswagen-Uber alliance each staking their claim.
Tesla confirmed its robotaxi service remains on track for a June/July launch in Austin, representing Elon Musk's most concrete timeline yet for a service that has seen numerous delays.
A significant development occurred when Tesla announced on X that it's already testing a supervised version of its upcoming ride-hailing service among employees in the Bay Area and Austin. This internal testing program has reportedly completed more than 1,500 trips and 15,000 miles of driving, serving as a precursor to the planned public rollout of unsupervised driverless robotaxis in two months.
While the initial deployment of just 10-20 Model Y vehicles (not the futuristic Cybercabs) seems modest, it signals at least Tesla's intent to compete directly with Waymo in Austin. These vehicles will have remote operators as backup, with plans to expand to additional US cities by year-end.
As industry expert
"If the launch features vehicles with no safety drivers, I will be impressed. But the other thing I'm watching is who will own the vehicles and shoulder the liability."
This is a crucial distinction from Waymo's approach, where safety is consistently emphasized as paramount.
As Campbell points out, "Elon's style" has typically involved pushing boundaries and challenging conventions, which has brought him considerable success, "but there's no getting around the lawsuits and liability that come with owning a fleet of shared vehicles."
The true test of Tesla's Full Self-Driving technology will come if they both remove the driver and accept fleet ownership liability.
Musk has already indicated that unsupervised FSD will come to personal Tesla vehicles "before the end of this year," suggesting he sees the transition from robotaxis to consumer vehicles as relatively straightforward. He even claimed Tesla vehicles are already driving themselves autonomously from the factory to parking lots.
However, it's important to recognize that these routes are pre-mapped, low-traffic, and extremely short—just 1.4 miles for the Model Y and 0.6 miles (mostly in an underground tunnel) for the Cybertruck in Texas—a far cry from the complex environments a true autonomous vehicle must navigate.
Musk's characteristic confidence was on full display during Tesla's earnings call when he boldly predicted the company would capture "90-something percent" market share in autonomous driving, dismissing Waymo's current lead in the space.
"I don't see anyone being able to compete with Tesla at present," he declared, arguing that Waymo's lidar-equipped vehicles are too few and too expensive—punctuating his point with a dad joke: "The issue with Waymo's cars is it costs way-mo money."
When pressed for additional details about the robotaxi rollout, Musk simply responded, "It's only a couple of months away, so you can just see it for yourself in a couple of months in Austin."
Speaking of Waymo, Alphabet's autonomous driving unit announced this week that it now provides over 250,000 paid rides weekly across San Francisco, Los Angeles, Phoenix, and Austin – up from 200,000 in February. This impressive growth demonstrates increasing consumer adoption and validates Waymo's methodical approach to market expansion.
But perhaps more consequential were the strategic comments from Alphabet CEO Sundar Pichai during the company's first-quarter earnings call. Pichai elaborated that Waymo has options in terms of "business models across geographies," and explained that the robotaxi company is building partnerships with ride-hailing app Uber, automakers, and operations and maintenance businesses that tend to its vehicle fleets.
"We can't possibly do it all ourselves," Pichai candidly acknowledged. He further noted that Waymo "has not entirely defined its long-term business model," and specifically mentioned "future optionality around personal ownership" of vehicles equipped with Waymo's self-driving technology.
These comments represent another strong hint that consumer-owned autonomous vehicles might be part of Waymo's future strategy. While unit costs would need to decrease substantially to make this viable for the mass market, a licensing business model—where Waymo provides its autonomous driving technology to OEMs—could be an attractive extension of their current operations.
Similar licensing approaches are being pursued by competitors like Wayve (recently partnering with Nissan) and Nuro. Given Waymo's position as the clear market leader in self-driving technology and its increasingly impressive operational metrics, such a B2B technology licensing play would likely find receptive partners throughout the automotive industry, opening doors that might otherwise remain closed to newcomers in the space.
Interestingly, Grayson Brulte of the "Road to Autonomy" podcast spotted a potentially significant detail in Waymo's announcement tweet about their 250,000 weekly rides.
Waymo used the term "powered by our generalizable Waymo driver," which could indicate a fundamental technological shift.
The term "generalizable" might suggest that Waymo's system no longer relies as heavily on HD maps—a significant departure from their traditional approach and potentially moving closer to the end-to-end AI methods used by competitors like Tesla and Wayve. If accurate, this would represent a major strategic pivot for Waymo, though we need more clarity on exactly what Waymo means by "generalizable”. This terminology choice could signal Waymo's response to critiques that their map-dependent approach limits scalability compared to vision-based competitors.
Meanwhile, Uber's AV partnership team continues to do what it does best: deliver new autonomous driving collaborations. This week, they announced yet another strategic alliance, this time with Volkswagen and its autonomous mobility unit MOIA.
The companies plan to deploy thousands of electric ID Buzz vans in the US, with testing set to begin later this year and a commercial launch scheduled for 2026 in Los Angeles. During the initial testing phase, the vehicles will operate with human safety operators on board, and the partnership will only proceed after securing necessary regulatory approvals. The companies have indicated plans to gradually expand to other major US markets over the next decade.
In a conversation with people familiar with the announcement, I gained deeper insights into what this partnership means for both companies. Volkswagen's contribution extends far beyond simply providing vehicles equipped with self-driving technology.
The German automaker is delivering a comprehensive ecosystem that includes not only the autonomous vehicles themselves, but also sophisticated fleet management software designed to optimize utilization and maximize operational efficiency.
This includes their Self-Driving stack integrated with Mobileye, the specially developed ID.Buzz vehicle with a robotaxi-optimized interior, comprehensive AV ecosystem software and operational enablement.
This seems to be a good match. Uber is a master in demand generation, matching, and balancing. Meanwhile, Volkswagen brings the ready-to-deploy solution that allows AVs to be integrated into Uber's platform almost as easily as human gig drivers. This partnership structure plays to each company's strengths while minimizing their respective weaknesses.
I find it particularly exciting to see a European manufacturer like Volkswagen entering the US market with their autonomous capabilities. The competitive landscape is diversifying rapidly, and the industry appears to be approaching a crucial inflection point where multiple business models will be tested simultaneously in the market.
Regulatory developments are also accelerating in parallel with these industry moves, potentially reshaping the competitive environment. Let's examine these changes in more detail.
US Regulatory Environment: Trump Administration Unveils New AV Framework
The regulatory landscape for autonomous vehicles in the US received a significant update this week as Transportation Secretary Sean P. Duffy unveiled NHTSA's new Automated Vehicle (AV) Framework. This initiative, part of the administration's broader transportation innovation agenda, aims to "unleash American ingenuity, maintain key safety standards, and prevent a harmful patchwork of state laws and regulations."
Duffy framed the changes in geopolitical terms, stating, "This Administration understands that we're in a race with China to out-innovate, and the stakes couldn't be higher. As part of DOT's innovation agenda, our new framework will slash red tape and move us closer to a single national standard that spurs innovation and prioritizes safety."
The framework is built on three core principles: prioritizing safety of ongoing AV operations, removing unnecessary regulatory barriers to innovation, and enabling commercial deployment of AVs to enhance safety and mobility. These principles are being translated into concrete policy changes that could accelerate AV deployment across the country.
One of the most significant changes is the expansion of the Automated Vehicle Exemption Program (AVEP) to include domestically produced vehicles. Previously, this program—which allows companies to operate non-compliant vehicles on U.S. roads—was only available for imported vehicles. By extending these exemptions to American-built autonomous vehicles, the administration is removing what NHTSA Chief Counsel Peter Simshauser called a "needless roadblock to innovation."
The changes also include modifications to NHTSA's Standing General Order on Crash Reporting for vehicles equipped with advanced driver assistance systems and automated driving systems. While maintaining the reporting requirement, the agency is streamlining the process to "sharpen the focus on critical safety information while removing unnecessary and duplicative requirements."
Meanwhile, California's Department of Motor Vehicles announced on Friday that it is proposing to allow testing of self-driving heavy-duty trucks and other large vehicles on state public roads. The proposal includes significant safety requirements—manufacturers would need to conduct at least 500,000 autonomous miles of testing (up to 400,000 of which may occur outside California) before applying for a driverless testing permit for heavy-duty vehicles. The state is also proposing expanded reporting requirements, including monthly data on disengagements, vehicle immobilizations, and hard braking events.
These concurrent regulatory developments at both federal and state levels demonstrate the increasing governmental focus on creating frameworks that balance innovation with safety concerns. For companies like Tesla, Waymo, and others, a more streamlined regulatory environment could significantly accelerate the timeline for commercial deployment of autonomous vehicles across the United States.
🇨🇳 China: Pony.ai Advances Technology While Navigating Geopolitical Challenges
In China, Pony.ai is making remarkable technological progress while navigating an increasingly complex geopolitical environment. At the Shanghai Auto Show, the company unveiled its seventh-generation autonomous driving system, showcasing dramatic cost reductions of 70% in its autonomous driving kit – including an 80% decrease in computing costs and 68% reduction in LiDAR expenses. These advancements address one of the most significant barriers to large-scale deployment and could accelerate Pony.ai's path to profitability compared to its Western competitors.
The cost efficiencies are complemented by Pony.ai's operational metrics – their Robotaxi fleet has completed 500,000 hours of all-weather, driverless operation with safety rates allegedly 10 times higher than human drivers. The company also revealed Robotaxi models developed in partnership with Toyota, BAIC, and GAC, demonstrating its ability to work with both international and domestic automakers.
Further strengthening Pony.ai's position is a newly announced strategic partnership with Tencent that will integrate its Robotaxi services into WeChat's "Mobility Services" and Tencent Maps. This creates a powerful ecosystem advantage by leveraging China's super-app infrastructure to reduce customer acquisition costs and create a seamless user experience – a model that Western companies may find difficult to replicate due to their more fragmented digital ecosystems.
Intriguingly, despite rising US-China trade tensions, Pony.ai appears relatively unfazed by the potential impacts of tariffs on their operations. Speaking on the sidelines of the Shanghai Auto Show, CEO James Peng told Reuters that the company does not foresee direct impacts from the Trump administration's tariffs on their supply chain.
"On the supply side, we always try to have alternatives and backups," Peng explained, highlighting their resilient supply chain strategy. The company currently uses Nvidia's autonomous-grade Orin-X chips and plans to utilize the newer Drive Thor chips, which are not restricted from being sold in China. Peng seemed confident about continued access to these critical components, stating, "It's very hard to imagine those chips will be on restriction."
Nevertheless, he acknowledged that the company has prepared contingency plans, noting, "Alternatively, of course, we also have some backup suppliers from the Chinese domestic chip manufacturers." This dual-sourcing approach appears designed to insulate Pony.ai from potential future restrictions.
While Peng admitted that the geopolitical situation would affect them "sentimentally," particularly regarding their international expansion plans, his primary focus remains on scaling production.
"Currently all the assembly line has been modified for the mass production. Our target is that for this year we'll reach close to 1,000 vehicles," he said. Looking further ahead, he projected even more ambitious growth: "For the next few years, we're going to actually very rapidly expand the offering and mass production once we have the assembly line all tooled up. I think now is the matter of the demand. We definitely are going to go beyond tens of thousands very soon."
Pony.ai appears to be getting serious about scaling. The CEO has repeatedly emphasized plans to reach 1,000 vehicles this year and expand to 10,000 vehicles in the near future. I'm curious to see if we'll notice the first signs of accelerating revenue in their next quarterly earnings report. So far, most of their revenue has come from their robotruck business, which doesn't generate as many headlines as their robotaxi operations. It will be fascinating to watch how this unfolds.
🔗 Reuters (1)/ Reuters (2) / Reuters (3) / Reuters (4) / CNBC / Sherwood (1) / Sherwood (2) / Sherwood (3) / Financial Times / U.S. DOT / Bloomberg (1) / Bloomberg (2) / TechCrunch / Gasgoo (1) / Gasgoo (2) / Pony.ai / Road to Autonomy
💡 Quick Takes:
🇳🇴 Oslo Launches Autonomous Vehicle Public Transport Trial
Norway has deployed five self-driving electric vehicles in Groruddalen, Oslo, marking the first of three planned launches in the EU-funded ULTIMO initiative. The project aims to establish the framework for large-scale, on-demand autonomous public transport services. Initial user feedback has been positive, though researchers acknowledge scaling will require approximately 500 vehicles to meaningfully impact private car use in the area. This initiative demonstrates how autonomous technology can address public transportation challenges in urban environments.
🔗 Horizon
🇯🇵 Nippon Steel and TIER IV Partner on Autonomous Heavy-Duty Vehicles
Japan's largest steel manufacturer, Nippon Steel, is collaborating with autonomous driving software developer TIER IV to automate steel transportation at its Nagoya plant by fiscal year 2025. The partnership addresses labor shortages by developing autonomous systems for specialized transporters that carry steel plates. This industrial application highlights how autonomous technology extends beyond passenger vehicles to solve specific logistics challenges in manufacturing environments.
🔗 Quality Digest
🇬🇧 UK-Based Wayve Expands to Japan
AI startup Wayve Technologies has opened a test center in Yokohama, making Japan its fourth market after the UK, US, and Germany. The SoftBank-backed company recently announced a partnership with Nissan to enhance its ProPilot driver-assist system for 2027 release. Wayve's CEO Alex Kendall emphasized their hardware-agnostic approach and data collection capabilities as competitive advantages, though regulatory hurdles in Japan remain significant. This expansion shows how AI-first approaches to autonomy are gaining traction globally.
🔗 Bloomberg / LinkedIN
📚 Worth Reading
Autonomous Vehicles: Timeline and Roadmap Ahead
This comprehensive WEF report provides a grounded perspective on the adoption timeline for autonomous vehicles between 2025-2035, covering personal vehicles, robotaxis, and autonomous trucks.
🔗 World Economic Forum
📊 Weekly Performance
Note: Stock performance data as of April 27, 2025. Past performance does not indicate future returns.
Thanks for reading!
If you found value in this newsletter, please consider sharing it with a friend or colleague who might benefit from these insights.
And if you haven't already, subscribe to stay updated on the latest developments in the autonomous driving industry.
Thanks for the shout-out! The big thing with Tesla is that their fans are such sycophants and their haters are such haters, it's hard to gauge where the actual tech is at. And I'm also no tech expert, so I prefer to look at more binary metrics like who's paying for insurance/owning the cars/etc as a starting point for how safe the FSD tech is :)