Waymo Expands to Three New Cities While Fixing Water Detection Problems

Waymo, a self-driving car company owned by Alphabet, has launched autonomous ride-hailing services in Miami, Orlando, and Nashville, according to an April announcement from the company. This expansion marks a major step forward—but Waymo is also dealing with a serious technical problem that forced it to recall 3,800 vehicles and temporarily shut down service in San Antonio.

The three new cities represent Waymo's biggest geographic push yet, bringing driverless taxi service to Florida and Tennessee for the first time. The rollout follows a careful approach: starting with human safety drivers on board, gradually removing the drivers as the system proves itself, and working with existing ride-hailing platforms.

The Water Detection Problem

While expanding into new markets, Waymo has faced a significant challenge with its vehicle sensors. The company recalled 3,800 robotaxis after discovering they couldn't reliably measure how deep standing water was—the kind of puddles or flooded areas that form after rain or near waterways.

This became urgent after two incidents in San Antonio. In one case, a Waymo vehicle was swept away by floodwaters. In another incident in May, a passenger had to get out of a vehicle that drove into flooded water. These events led Waymo to pause San Antonio operations while it works on a fix.

This problem highlights something fundamental about self-driving cars: humans learn intuitively how to handle unusual situations, but machines need explicit programming for everything. Detecting water depth is particularly tricky because cameras and lidar sensors—the equipment that lets the car "see"—struggle to measure how deep water is, especially in poor light or bad weather.

The broader context here is that water-related incidents matter. When a human driver sees water across a road, they know to slow down or stop. A self-driving car without robust water detection can end up in genuinely dangerous situations, and that's exactly what happened.

Partnering With Waze for Hazard Detection

Waymo has announced a partnership with Waze, Google's navigation app, to identify and report hazards—starting with potholes. The idea is straightforward: Waymo's fleet of robotaxis carries high-quality sensors that can spot problems in the road, and that data feeds into Waze so human drivers can see it too.

This is part of a wider industry trend where autonomous vehicle fleets do double duty: they provide transportation and they gather detailed information about road conditions. Waze gets better data from Waymo's professional-grade sensors, and Waymo gets access to information from Waze's millions of users about where hazards are.

Different Vehicle Types and Testing

Waymo is also diversifying the types of vehicles it uses. Along with its current fleet of modified Chrysler Pacificas and Jaguar I-PACEs, Waymo is now adding Hyundai electric vehicles and testing purpose-built robotaxis made by Zeekr in San Diego. Purpose-built means these vehicles were designed from the ground up for autonomous operation, rather than being converted from cars originally made for human drivers.

Using different vehicle types gives Waymo flexibility: it reduces dependence on any single supplier, allows the company to optimize costs, and lets Waymo test which vehicle designs work best in different situations. Purpose-built robotaxis typically offer more interior space, easier maintenance access, and better integration of sensors compared to converted consumer vehicles.

When the System Breaks Down

Beyond water detection, Waymo has run into reliability problems that expose how complex it is to keep a fleet of driverless cars running smoothly. At one point, robotaxis stalled mid-route and lost connection to the central systems that coordinate them, causing traffic disruptions.

This reveals an important vulnerability: unlike human-driven cars that operate independently, robotaxis depend on constant communication with cloud-based systems that handle routing, monitoring, and safety overrides. If the network fails, the servers crash, or software has a bug, hundreds of vehicles can be affected simultaneously.

Having watched cloud computing evolve over the past two decades, I recognize this pattern. When cloud services first rolled out, companies discovered the same challenge: you gain intelligence and flexibility by centralizing control, but you also create a potential single point of failure. The industry eventually learned to build redundancy into these systems—backup servers, distributed processing, the ability to fall back to local operation. Waymo will likely need to do the same.

Competition Heating Up

Waymo's expansion is happening in a crowded market. Lyft, the major ride-hailing company, recently partnered with May Mobility to launch autonomous taxi service in Atlanta. Instead of building their own self-driving technology, established companies like Lyft are partnering with specialized autonomous vehicle makers.

This partnership model may end up being more practical than trying to do everything in-house. Ride-hailing companies focus on managing demand, setting prices, and keeping customers happy, while the autonomous vehicle specialists handle the difficult work of making cars drive themselves safely.

What Comes Next

Waymo's situation right now captures where the autonomous vehicle industry actually is: capable in controlled environments, but still running into problems that reveal system limitations. Water detection is one category of edge cases—situations that are uncommon but serious when they occur—that the technology still needs to handle better.

In my view, the partnership with Waze is worth watching. It suggests a future where robotaxi fleets serve as distributed sensing networks, constantly gathering and reporting data about road conditions. If this works, it could make the business case for self-driving taxis stronger by creating additional revenue from the data itself, not just from rides.

How Waymo performs in Miami, Orlando, and Nashville will matter. The company needs to manage the complexity of running operations across multiple cities while also solving the technical problems that have emerged. That Waymo recalled vehicles and paused San Antonio service rather than pushing through suggests it takes safety seriously—an approach that builds confidence in the long term, even if it slows short-term expansion.