Mercedes-Benz will put Wayve's Level 2 driver assist in a model due within two years

Sometime in the next two years, someone will buy a Mercedes-Benz with Wayve's software inside and still be the one driving it ^[1]. The system is Level 2. It handles certain driving maneuvers, and the human driver has to remain engaged ^[2].

What Wayve is pitching is a fully driverless product ^[4]. What it is doing at Mercedes is driver assistance, sold through a carmaker. Kirsten Korosec wrote in TechCrunch Mobility: "While this is not a Level 4, or fully driverless product, it gives Wayve reach and follows similar deals with Nissan and Stellantis." ^[3] That makes Mercedes the third carmaker named with a deal of this kind ^[3].

She also wrote: "Those partnerships have also opened doors for its fully driverless product." ^[4] The example on record is a robotaxi service in Tokyo that Wayve announced earlier this year with Nissan and Uber ^[5]. Nissan is on both lists, as a Level 2 customer and as a robotaxi partner ^[3][5]. TechCrunch did not report deal terms or name the Mercedes model. So the reporting cannot show whether assist revenue pays for the driverless program or whether the deals mostly lead to introductions like the Tokyo one.

Waymo scales with its own fleet and keeps that fleet concentrated. Korosec's review of vehicle registration data put about 80% of its roughly 4,000 robotaxis in California and Texas ^[6]. That is about 3,200 cars in two states ^[1]. The Texas fleet grew more than 49% in three weeks ^[7], and Waymo is now going after teenagers as riders ^[8]. In a pitch deck, both approaches count as autonomous driving. Here's what users actually do: a Waymo rider is a passenger, and a Wayve user at Mercedes is still the driver ^[2].

Aurora, which builds self-driving trucks, is focused on Texas but has cast a wide net for partners, Korosec wrote ^[9]. Chief executive Chris Urmson said this week that Aurora has "emerged from the building stage" ^[10]. The company plans to have more than 200 driverless trucks by the end of this year and more than 30,000 in operation by 2030 ^[12][11]. Hitting that target means growing the fleet about 150-fold in roughly four years ^[2].

I'd sort any deal with these companies on two axes. The first is who drives: a human who stays engaged, or nobody. The second is who supplies the vehicles: a partner, or the AV company's own fleet. Wayve's deals with Mercedes, Nissan and Stellantis have a human driving and a partner supplying the car ^[2][3]. Wayve's Tokyo plan is driverless and built with partners ^[5]. Waymo runs driverless cars on its own fleet ^[6], and Korosec puts Aurora between the two camps ^[9]. Which box a deal sits in decides the number to ask for. For a Level 2 deal, ask for a named model with a ship date, and later for the share of drivers still using the feature months after they buy the car. For a driverless deal, ask how many vehicles are in operation and where they are. Waymo's 4,000 look different once you know 80% of them are in two states ^[6].