Aurora Innovation Pushes Back Gross-Margin Timeline to 2030 While Unveiling Asset-Light Hardware Deal With AUMOVIO, Even as Key Customer Admits an Economics "Gap" Still Exists
Analyst and Investor Day, Dallas, September 23, 2026
Aurora Innovation used its first investor day in two and a half years to lay out a long-term financial roadmap that is both more ambitious and, in places, more conservative than what the market had been modeling. The autonomous trucking company now guides to $5 billion in revenue and more than 30,000 trucks on the road by 2030, but it also pushed back its positive gross-margin target and trimmed its long-run gross margin assumption from 70% to "above 60%," a subtle but meaningful downward revision that CFO David Maday attributed to supply chain headwinds and a desire to be "appropriately conservative."
Gross profit and free cash flow targets slip
The most consequential disclosure of the day was timing. Aurora now expects to hit positive gross profit on a run-rate basis in the first half of 2027, a change from prior guidance of exiting 2026. Maday was direct about the cause: "This target slightly moved from where we were talking about the end of this year, and that's really just reflecting a slightly slower fleet ramp." Positive free cash flow on a run-rate basis is now targeted for 2028, at roughly 7,500 trucks in operation. The company still expects to exit 2026 with 200 driverless trucks, translating to an $80 million revenue run rate on its owned-and-operated Transportation-as-a-Service fleet, and to end 2027 with more than 1,000 trucks generating roughly $200 million in revenue.
AUMOVIO deal reshapes the capital model
The most structurally important new information was the depth of Aurora's third-generation hardware partnership with AUMOVIO, the Continental spinoff. The two companies are building what Aurora calls an industry-first Hardware-as-a-Service structure, where AUMOVIO's hardware costs are paid per mile rather than upfront, eliminating capital expense for carriers and allowing Aurora to run an asset-light model consistent with software-like margins. Jeremy McClain of AUMOVIO called it a deliberate bet on a new business model: "It's all about aligned incentives... we get incentivized to design it properly and as well to make sure that it operates properly in exactly that way." Start of production is slated for the second half of 2027, with material economic benefit expected in 2028. Aurora's own capital expenditure guidance reflects the transition: $185 million in 2027, mostly tied to buying out the remaining second-generation hardware kits and Transportation-as-a-Service trucks, dropping below $50 million in 2028 and to under 1% of revenue longer term.
A customer admits the economics still don't fully pencil
Perhaps the most candid moment of the day came not from Aurora but from Werner's Chief Information Officer, Daragh Mahon, who told the room bluntly that the pricing conversation with Aurora is unresolved. "I will be quite honest that we have a gap. We've got to figure this out," Mahon said, adding that he expects the economics to become "really viable at scale" within "the next few months." That admission tempers the otherwise-triumphant tone of the day and suggests the $0.85-plus per mile Driver-as-a-Service pricing Aurora is targeting for 2027 may need further negotiation before major carriers commit at volume. Mahon and McLane's Eric Hildenbrand were both clear that long-haul, higher-mileage lanes generate the best economics, while shorter or more variable routes remain harder to underwrite financially even where the safety case is strong.
Uber's stake sale resolves an overhang
Chris Urmson addressed Uber's recent share sale directly, characterizing it as a positive rather than a warning sign. Uber, he said, has been reallocating capital to fund competitive positioning in robotaxis, a move Aurora "saw coming." He argued the block of stock had been "a large, concentrated position from a holder that had no intent to hold it long term," and that its recirculation to longer-term holders removes a stock overhang rather than signaling any deterioration in the commercial relationship.
Insurance economics get real numbers attached for the first time
In a rare disclosure for the sector, Aurora brought Chris Moore, Chief Underwriting Officer at Apollo ibott (a Lloyd's of London syndicate), on stage to discuss how autonomous truck risk is actually being priced. Moore said Aurora's current premiums run "slightly above" those for human drivers, driven by uncertainty around legal severity rather than frequency of accidents, but that rates are falling 15% to 25% annually as claims data accumulates. He was unusually specific about the mechanics of liability defense in court, noting that Uber's mixed human/autonomous fleet has inadvertently given insurers a defensible baseline: "Now they can have a direct comparison to say that the autonomous vehicles operating are 90% less likely to get into an accident... I've actually connected you with the safest form of transportation. You can't come for me." Moore also flagged that few competitors are entering the space because incumbent insurers view autonomy as cannibalizing their existing loss-data-driven business models, a dynamic that is keeping capacity tight even as Aurora's own risk profile improves.
Lane expansion economics are inflecting faster than expected
Urmson offered a data point on operational velocity that underpins the company's scaling thesis: the first commercial lane took six years to develop, validate and deploy; the second took six months; a more recent one took six weeks; and management now expects future lanes to take days. He described this as more than linear improvement, "not just a linear improvement in how quickly we can open lanes, but an exponential one," driven by increasing generalization of the underlying driving system. Management also detailed unit economics on a specific Phoenix-to-Fort Worth lane showing a doubling of revenue and a sixfold increase in margin per truck annually once autonomy is deployed, equivalent to $340,000 in incremental revenue and $160,000 in incremental margin per truck per year.
Multi-OEM hardware strategy aims to de-risk supply
Aurora and its manufacturing partners spent considerable time addressing single-source risk, a live concern among analysts given the capital intensity of scaling. PACCAR's Noelle Onstad said the company is building its autonomy-ready trucks on a shared "AVP" platform designed to accommodate multiple technology partners in the future, while Volvo Autonomous Solutions' Sasko Cuklev said Volvo's CAST architecture already supports multiple partners across its truck brands, including Mack and Renault. Volvo confirmed it expects driverless operations in the first quarter of 2027 and plans to have more than 300 driverless trucks in service by year-end, with one large fleet customer telling Volvo it wants half of that allocation immediately, according to Cuklev.
Market sizing and long-term ambition beyond trucking
Management framed the U.S. opportunity as a $1 trillion market based on 200 billion vehicle miles traveled annually, with a near-term addressable segment of 60 billion vehicle miles by 2028, up from roughly 4 billion today. Even single-digit market share within that segment would represent more than $1 billion in revenue, according to Maday. Longer term, Urmson pointed to international expansion in Japan, Korea and the Gulf region, as well as adjacent applications in smaller trucks, off-highway mining and eventually robotaxis, though he was careful to note China remains "basically off limits" and that robotaxi economics are considered harder and more capital intensive than trucking. He was unambiguous that trucking remains the singular near-term priority: "There's no effort to take the foot off the gas in any way in that space... shame on me and shame on us if we let that slip."