Aurora promises 30,000 driverless trucks by 2030. Uber already sold $1.1 billion of that story

Twenty trucks now, thirty thousand promised

Drive I-45 between Dallas and Houston and you may pass an Aurora Innovation semi with nobody in the driver's seat; about 20 of them haul freight commercially today ¹. On 23 September, at its analyst day, the company promised more than 30,000 driverless trucks in operation by 2030 ². The answer came the next morning: the stock fell 8.78%↗ ³.

The 2030 fleet runs on other people's balance sheets

The plan only works if someone else buys the trucks. Aurora caps the fleet it owns and operates, Transport-as-a-Service, at roughly 500 vehicles; from 2027, growth shifts to Driver-as-a-Service, where carriers own the truck and pay Aurora $0.85 or more a mile, against roughly $2 a mile for Aurora-run trucks ⁴. The flagship commitment behind that shift, Hirschbach's 500 trucks from 2027, is a non-binding memorandum of understanding in Aurora's own disclosure↗ ².

Two million in, 270 million out

The second quarter brought $2 million of revenue against a $270 million net loss, with nearly $1.2 billion of cash and short-term investments left ⁵⁶. Full-year guidance↗ is $14 million to $16 million ⁶; even the 200-truck exit for 2026 annualizes to about $80 million of revenue ⁴, less than one quarter of losses at the current pace. Aurora has said it may need another $650 million to $850 million before free cash flow turns positive in 2028 ⁷.

Uber, its largest shareholder, has sold three blocks of stock since June for about $1.1 billion↗; the last, 29.4 million shares at $6.21, cleared on 15 September, eight days before the investor day, leaving about 157 million shares ⁶⁸. The selling began before the promise, not after it: the clearest read so far on the 1,500x pledge came from inside Aurora's own shareholder register.

The bull case is real, and already traded

Bank of America pegs Aurora's service near $0.85 a mile against $1.30 for a human driver's wages and benefits; Morgan Stanley argues a scaled autonomous fleet could be 7.5 times as profitable, an $18 target against Evercore's $8 ¹⁷. The spread names the trade: a robot mile at $0.85 leaves roughly $0.45 for the carrier or, once rates compete, the shipper; the carrier puts the truck on its balance sheet, and Aurora takes software-style margins on trucks it does not have to buy.

Aurora's own math claims customer savings above 20%, about $0.40 a mile ⁴. Those projections rest on 20 trucks, 500,000 driverless miles, and utilization above 225,000 miles a year against an industry average of 85,991 ²⁶. Chief executive Chris Urmson's rebuttal: "We've been able to put out timelines and we've basically hit those timelines" ¹.

Two signatures settle it

Upfitter Roush is due to reach 20 trucks a week in October ², leaving roughly 180 to add in the fourth quarter to hit 200 ⁶. Two signatures then settle it: whether Hirschbach's memorandum becomes a binding order, and whether Uber's remaining 157 million shares stay put ⁸. Aurora has automated the driving. It has not automated the funding.