Aurora Innovation Launches Driver-Out Trucks and Promises 50% Hardware Cost Cut, But Quietly Trims Truck Count Target From "More Than 200" to Just 200
Q2 2026 earnings call, July 29, 2026
Aurora Innovation used its second-quarter call to declare the start of its "commercial scaling phase," anchored by the debut of Aurora Driver 2, the company's second-generation hardware and software stack now running in trucks with no one behind the wheel. CEO Chris Urmson called the July product launch a milestone that "moves Aurora into the start of our commercial scaling phase," and the company says it is fully allocated to end 2026 with 200 driverless trucks on the road. Revenue guidance held steady at $14 million to $16 million for the year, up 400% year-over-year at the midpoint, with the fourth quarter expected to carry more than half of that total.
The hardware cost story investors actually care about
The most consequential disclosure was economic, not technical. Aurora says its second-generation commercial hardware kit, engineered for 1 million miles of operation, will drive "a 50% plus reduction in Aurora Driver hardware costs," which management frames as the key lever to reach breakeven gross margin. The kit also extends the range of Aurora's proprietary FirstLight FMCW lidar to 1 kilometer, giving the system more than 34 seconds of reaction time at highway speeds. CFO David Maday told analysts the company already has real cost data because it is building the kit today, not modeling a future one: "We have costs for the ones we're producing today and the ones that we'll be producing into next year... some minor increases in component costs... are not materially going to impact our gross margin projections." A third-generation kit, built with AUMOVIO and aimed at supporting "tens of thousands of trucks" starting in the second half of 2027, is described as the next step-function cost reduction needed to hit Aurora's ultimate margin targets.
A subtle walk-back on truck count
Bernstein's Justine Laufer, on behalf of David Vernon, flagged that Aurora's language shifted from "more than 200" driverless trucks by year-end last quarter to simply "200" this quarter. Maday downplayed the change, saying the company is "fully allocated" to 200 but retains flexibility to go higher, and that the gating factor is matching truck supply to confirmed customer demand rather than any softening in build rates: "We'll actually have more trucks built than that, but we also have to have some trucks that are able to support development and then we have to have trucks that can backfill during regular scheduled repair and maintenance." Still, the specificity of the original question suggests the Street is watching Aurora's truck count guidance closely as the clearest proxy for execution risk, and any further downward drift would be a warning sign given how central fleet scale is to the 2027 financial model.
Volvo's $3 billion bet and the economics behind it
Aurora disclosed that Volvo Autonomous Solutions plans to begin driverless operations on the Volvo VNL Autonomous in the first quarter of 2027, exit that year with more than 300 driverless trucks, and is targeting $3 billion in autonomous revenue within five years, a figure Volvo announced at its own Capital Markets Day. Urmson would not detail Volvo's underlying unit economics but confirmed Aurora has "a clear understanding of the economic arrangement between Aurora and Volvo." Notably, management characterized Volvo Autonomous Solutions not as an OEM integration partner alone but as a Driver-as-a-Service customer of Aurora's, a distinction that matters for how investors should model long-term revenue mix between OEM-run fleets and direct-to-carrier deals. PACCAR is also in the mix, jointly defining a path for Aurora's third-generation kit on its future autonomy-enabled platform.
TaaS-to-DaaS economics finally quantified
For the first time, management gave explicit per-mile revenue figures for its two business models: Transportation-as-a-Service revenue runs in the "$2-plus" per mile range, while Driver-as-a-Service is targeted at "$0.85 plus." That roughly 2.5x revenue gap comes with an offsetting cost and margin difference, since TaaS requires Aurora to operate the truck and carry the freight relationship, while DaaS simply licenses the driver system to asset owners like Hirschbach, Werner, or Schneider. Aurora reiterated that its Hirschbach MOU calls for 500 tractors to be added to that carrier's fleet across 2027 and 2028, and that every current TaaS customer is expected to migrate toward a DaaS contract in 2027, transforming the revenue mix over the course of that year rather than in a single cutover.
New customer wins build out the freight network
Aurora added Charger Logistics and Value Truck as new Transportation-as-a-Service customers, both launching from Laredo, Texas, a corridor management repeatedly cited as benefiting from nearshoring-driven freight volume. The company also expanded its Volvo-powered commercial freight service to DSV and AVI-SPL, and highlighted its Detmar Logistics deployment in West Texas, where driverless trucks now run between a Midland facility and a Capital Sand mining site along I-20, including validation of a frac sand trailer with what Aurora described as "minimal integration work." Management framed each new customer as a "pipeline multiplier" that shortens onboarding cycles for the next one, though onboarding speed gains appear to be as much a function of Aurora's own operating maturity as customer demand, per Maday's comment that adding a new route once took weeks even after the system was technically capable.
A safety anecdote management chose to lead with
Urmson spent unusual time on a single incident: an Aurora truck in manual mode, driven by a human operator with more than 2 million miles of Class 8 experience, was struck by a vehicle that ran a red light in Fort Worth. No serious injuries occurred, but simulation and log review showed the Aurora Driver software, which was not engaged at the time, would have detected the red-light runner nearly six seconds before impact and slowed to avoid the collision despite having the right of way. The story is clearly being used to reinforce Aurora's safety differentiation ahead of wider driverless deployment, and it lands at a moment when the company is asking regulators and freight customers to trust a system operating without any human fallback.
Regulatory tailwinds building, but not yet secured
California has joined the majority of states permitting driverless truck deployment, and Aurora has submitted its application to begin required driver testing there, with Urmson noting he personally testified before the California Assembly Transportation Committee. At the federal level, the House Transportation Infrastructure Committee passed the BUILD America 250 Act by a 62-2 vote, which would create a national framework for autonomous truck deployment and permit cab-mounted warning beacons in place of manually placed roadside devices, a change Aurora has pushed for. The legislation is not yet law, and management was careful to frame this as momentum rather than a completed win.
Balance sheet and cash burn
Aurora ended the quarter with roughly $1.2 billion in cash and short-term investments after raising $215 million in net proceeds through its at-the-market program, issuing 30 million Class A shares. Operating cash use was approximately $225 million for the quarter, including $63 million in cash bonus payments funded through the ATM. Full-year 2026 capital expenditures are still guided to roughly $150 million, with quarterly cash use expected to average $190 million to $220 million. Second-quarter revenue was just $2 million, and operating losses including stock-based compensation totaled $266 million, underscoring that Aurora remains firmly in a pre-commercial cash-burn phase even as management talks up an inflection point. The company reiterated it could support up to 500 test trucks if financing needs required it, suggesting balance sheet flexibility is not viewed internally as a near-term constraint.