News

Testing Chinese Robotaxis… Again // The Ride AI Summit Is Tomorrow

Plus, Waymo is heading to DC in 2026.

The big day is almost here: Our first-ever Ride AI summit is happening tomorrow (4/2) at NeueHouse Hollywood.

This one-day event will bring together top minds in autonomous vehicles and AI mobility for real talk on what’s happening now: breakthroughs, roadblocks, and strategies shaping the future of autonomy. We’ve curated a dynamic lineup of speakers and sessions—check out the full schedule here.

If you want tickets, please book soon—there are fewer than 10 spots left

Get Your Ride AI Tickets

Now, Here’s What’s Happening

Last week, I traveled to Wuhan, China, to try Baidu’s 6th generation Apollo Go vehicles firsthand. I came away extremely impressed, not only with the vehicle itself, but with the operations and infrastructure behind it. While the $30,000 BOM cost was headline grabbing, what I think is perhaps more important is the how the company recharges the vehicles. I highlight this in my video below.

The 6th generation Apollo Go vehicles use battery swapping rather than traditional charging. This process is completely automated with no human operators needed, taking only about 1.5-2 minutes per vehicle. After the swap, vehicles can immediately return to service.

With this model, Baidu could potentially eliminate the need for massive urban depots entirely. These stations could do double duty as data hubs, offloading gathered information via physical media swaps instead of wireless transfers. Vehicle cleaning—a less frequent necessity than charging—could be handled at dedicated facilities, creating a distributed network of specialized service points. While this model doesn’t yet account for demand curves and deadheading challenges, the efficiency gains are undeniable. I wonder which autonomy provider will be next to implement this solution…

Caterpillar announced a deal to use Luminar’s LiDAR sensors for self-driving off-road equipment. The OEM will start by integrating two of Luminar’s Iris sensors into vehicles working at quarry sites. According to Luminar, the Iris sensor is particularly effective in dusty environments, because it operates at a longer wavelength than most LiDAR sensors. Caterpillar is joining a number of other manufacturers using Luminar’s equipment, including Mobileye, Volvo, and Pony.ai.

We haven’t really seen LiDAR used in high-dirt, high-dust settings before, so I’m extremely curious about how the technology will fare. While I have no doubt that the sensor can see “through” the grit when processing is applied, I’m more interested in the hardware’s longevity. As a moving part, LiDAR is particularly susceptible to heavy and repeated vibration, impact, and particle ingress. I look forward to seeing how Luminar and Caterpillar solves these issues.

Pony.ai is now the first company authorized to operate fully driverless robotaxis in Shenzhen, beating Baidu’s Apollo Go to the punch. The new permit grants Pony.ai permission to charge customers for fully driverless rides in a service area that includes the Shenzhen Bao’an International Airport, Shenzhen Bay Port, and Nanshan district, where companies like Tencent and DJI are headquartered. While the permit allows for operation in “all weather conditions,” service is currently limited to the hours of 7:30AM to 10PM.

Apollo Go has historically been the poster child for robotaxis in China and has been testing in two separate service areas in Shenzhen, so it’s surprising to see the city award its first fully driverless permit to Pony.ai. (Side note: I’ll be on a panel about the rise of Chinese AVs with reps from both Apollo Go and Pony.ai tomorrow at Ride AI).

After suspending the sale of FSD, Tesla is dropping the “FSD” moniker entirely from its driver-assist features in China. The company is changing the name of the software from “FSD Intelligent Assisted Driving” (FSD 智能辅助驾驶) to “Intelligent Assisted Driving” (智能辅助驾驶). In addition, “Autopilot Automated Assisted Driving” (Autopilot自动辅助驾驶) has been renamed as simply “Assisted Driving” (辅助驾驶). This is not the first time Tesla has changed its branding. After initially launching FSD in China as “Full Self-Driving Capability” (完全自动驾驶能力) in February, Tesla quickly updated the nomenclature to “FSD Intelligent Assisted Driving.” It then paused the rollout of the feature set entirely in March, citing lack of regularly approval.

In Other News…

Alright, that’s it from me. Hope to see you at Ride AI tomorrow in Hollywood!

Join our weekly newsletter

Get all the latest news in AI x Mobility across the globe.