A Look At Waymo’s Rough Week
Plus, Baidu is not having a good time either.
Welcome to the Ride AI newsletter: your weekly digest of news and intelligence at the intersection of technology and transportation.
Ride AI 2026 will be held at SFJAZZ!
Due to an overwhelming response on the waitlist, we’ve decided to change venues. Ride AI 2026 will now be held at SFJAZZ, one of America’s most iconic stages. With a best-in-class audiovisual system and an intimate atmosphere that transforms presentations into experiences now in the cards, we’re prepared to go bigger. There are still some spots left on the waitlist, so make sure to get signed up to lock in early bird pricing.
Now, Here’s What You Need To Know Today.
The McKinsey x Ride AI Industry Survey Reveals a widening global gap.
The survey shows deployment timelines slipping 1 to 2 years across nearly every use case, with the global gap widening rather than narrowing. China remains far ahead, United States holds a strong second, and Europe continues to fall behind due to slower innovation cycles, fragmented regulations, and capital spread too thin across too many players. During our follow-up webinar, industry leaders from BMW, Mobileye, Torc Robotics, MOIA, and McKinsey emphasized that Europe’s strict homologation requirements push Level 4 timelines out by years, while the United States benefits from faster state level approval and more flexibility for commercial pilots.
Autonomous trucking appears closer to economic viability than urban robotaxis, driven by the United States’ eight hundred billion dollar freight market, extreme driver turnover, and simpler highway operational domains. Torc Robotics highlighted why long haul routes like Dallas to Laredo offer a clearer path to early commercial scale, especially with factory integrated autonomous ready chassis from Daimler Trucks. Meanwhile, the survey revealed cooling expectations for Level 3 in private vehicles as costs rise and timelines stretch, though BMW and Mobileye say driver demand for eyes off features remains strong.
Across the industry, experts overwhelmingly believe hybrid AI architectures will dominate rather than pure end to end systems, reflecting regulator demands for transparency and explainability. Companies are leaning into world simulation and reinforcement learning with guardrails to accelerate validation and expand operational domains safely. The long term outlook points to regional divergence, with China pushing ahead fastest, the US accelerating through commercial partnerships, and Europe moving more slowly until cost parity with human drivers arrives around 2030 to 2032.
The full webinar is now available for your viewing pleasure on the Ride AI YouTube channel. Let us know what you think of the results and datapoints shared, and if you like it, don’t forget to subscribe!
Waymo has had a rough past week as it continues building toward a global robotaxi rollout.
The company says it has gone driverless in Houston, begun testing with safety drivers in Philadelphia, and will start manual data-collection drives in Baltimore, St. Louis, and Pittsburgh. It also announced yesterday that the Waymo Driver had landed in London and that testing had begun. Waymo has not provided a timeline for commercial service in any of the new markets or indicated whether it plans to partner with other companies, as it has done with Uber in places like Atlanta and Austin, but the additions bring the total number of cities in Waymo’s pipeline to more than twenty, spanning active robotaxi service, freeway routing, commercial prep, and early-stage testing.
However, it’s not been all smooth sailing for Waymo. The NHTSA opened an investigation in October after video showed a Waymo vehicle maneuvering around a stopped school bus in Atlanta with its stop sign extended. Additional incidents were later reported in Austin, where school district officials allege the company’s vehicles passed buses illegally multiple times, including after Waymo said it had issued earlier fixes. Federal regulators have since requested detailed information about Waymo’s fifth-generation autonomous system and its operating procedures, which has led to the company filing a voluntary software recall with regulators after identifying issues with how its robotaxis behave around school buses.
The recall will be submitted early next week and covers scenarios in which autonomous vehicles must slow or stop for buses that are loading or unloading children. Waymo had updated its software on November 17 after identifying the problem and says the fix has already improved performance beyond typical human driver behavior.
That’s not all, however. Newly released security footage of a Waymo vehicle’s killing of a beloved neighborhood cat shows a slightly different account of events than what the company had released previously. The original report from Waymo had neglected to mention that a person had been crouched in front of the car, trying to coax out the cat that had crawled under it. The car had started moving right after the person had moved out of the way. Waymo explained this exclusion as an attempt to avoid potential blame the person could receive for being the cause of the cat’s death. However, it still means the company had left out important information about the sequence of events, and raises questions around why the vehicle moved so quickly and close to the crouching person.
News also broke on Sunday night of another Waymo vehicle hitting and killing an unleashed dog that had run out into the street. While no amount of caution could have avoided this unfortunate collision, the news became more ammunition for robotaxi detractors. In addition, a video posted to X on December 5th showed a Waymo vehicle driving through a flooded parking lot. Another video shows a Waymo vehicle driving right into the middle of a police standoff, and yet another video shows two Waymo vehicles colliding at low speed while a third Waymo vehicle is stopped by the ensuing traffic jam while heading the other way. This is not the first time a Waymo vehicle has driven straight into a flooded or police-active area, and while the company has publicly stated that it would work on improving the behavior, all these mistakes in such a short period of time has certainly done some damage to public perception around the company.
Waymo is not the only company having a bad week however…
A serious robotaxi accident in Zhuzhou, China leaves two pedestrians in intensive care.
Witnesses say the vehicle, operated by Hello (formerly Hellobike), struck two people around 9 AM on Yanjiang Road shortly after passing a pedestrian crossing. One victim became trapped beneath the car while another was injured nearby. Videos circulating on social media show bystanders attempting to lift the vehicle to free the trapped pedestrian, who appeared to be wearing a helmet and had visible facial injuries. Emergency responders transported both victims, one male and one female, to Hunan Provincial Hospital of Traditional Chinese Medicine, where they remain in the ICU. Authorities have not released details on the severity of their injuries.
Hello began autonomous road testing in August with a fleet of about thirty vehicles operating under regulatory approval. That regulatory approval was quickly rescinded in Zhuzhou after the incident. Although Hello recently unveiled its own L4 autonomous HR1 vehicle and promised mass production by June 2026, the robotaxi involved in the crash was a Baidu Apollo RT6 model purchased for early testing, running Hello’s own software. The RT6 is Baidu’s sixth-generation autonomous platform and has been widely used across Apollo Go’s deployments in cities such as Beijing, Shanghai, Shenzhen, Hong Kong, Dubai, and Abu Dhabi. Hello’s program was far smaller in scale, being limited to Zhuzhou and Liyang.
Neither the authorities nor the related companies have released any statements on the issue, so visibility into what really happened is unclear.
Uber and Avride have launched a commercial robotaxi service in Dallas.
The service begins with human safety operators and a nine-square-mile operating zone that includes downtown Dallas. Uber and Avride say fully driverless operations will come later, along with a broader service area. The fleet consists of all-electric Hyundai Ioniq 5 vehicles running Avride’s self-driving system, with Uber handling rider support from day one and planning to take over fleet operations as the service scales. Riders requesting UberX, Uber Comfort, or Uber Comfort Electric may be matched with a robotaxi at the same fare as a human-driven trip, and the Uber app will handle unlocking the car, opening the trunk, and starting the ride.
The launch caps a year of rapid AV expansion for Uber, which has now partnered with twenty autonomous vehicle companies across robotaxis, freight, and delivery. The company already offers robotaxi trips with Waymo in several U.S. cities and operates autonomous rides with WeRide in Abu Dhabi and Riyadh. Avride, backed by a recent three hundred seventy five million dollar investment involving Uber and Nebius Group, is also expanding its sidewalk delivery robots through Uber Eats. Uber expects autonomous vehicles to be available in at least ten cities on its network by the end of 2026, with upcoming rollouts planned in Arlington, Dubai, London, Los Angeles, Munich, and the Bay Area.
In Other News…
New ‘KnoWay’ robotaxis cause chaos in upcoming Grand Theft Auto Online DLC
Driverless trucks could soon be headed to California highways
Geely’s ride-hailing platform Caocao Mobility unveils ambitious global robotaxi strategy
New version of Tesla FSD allows texting while driving “Depending on context of surrounding traffic”
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!









This roundup does a great job showing how incident clustering damages public trust way faster than steady progress builds it. The cat situation is particularly telling, omitting the crouching person from the intial report feels like defensiveness more than transparency. Waymo's school bus issue showing up across multiple cities signals a pattern, not an edge case, and thats the kind of thing that regulators latch onto.