The driverless software gets the headline. The fleet work decides whether a rider gets picked up. Waymo and Uber’s September 2024 agreement put a specific split behind the robotaxi headline. Waymo supplied the automated driver. Uber would offer rides in its app and manage the fleet in Austin and Atlanta, including cleaning, repair, and depot work. The plan said the fleet could grow to hundreds of vehicles; that was a target, not the starting count.
I had focused on Waymo’s weekly trip number as if it proved the business case. Paid trips show that riders will use a service. They do not disclose what Uber pays to keep a vehicle clean, charged, available, and repaired, or what Waymo spends supporting the driverless system. The partnership is interesting precisely because it assigns that work instead of pretending the car runs itself.
Waymo now lists Austin and Atlanta rides through Uber, so the announcement became a real service. Its public page also says the fleet serves limited areas and favors requests near an available vehicle. That is the operating constraint I would keep in view when someone calls a city ‘launched.’
I have ridden Waymo in San Francisco. The ride convinced me the product can feel ordinary to a passenger. This deal is a better test of whether two companies can make the unglamorous fleet work ordinary too. The robotaxi map keeps the product and the labor behind it separate.