Aurora Innovation CEO Targets Hundreds of Driverless Trucks and $80M Run Rate by End of 2026

Aurora Innovation (NASDAQ:AUR) Chief Executive Officer Chris Urmson told investors at Morgan Stanley’s TMT conference that the company’s past year marked a major shift from development to real-world deployment, and that 2026 is expected to bring a significant increase in fleet scale, new hardware, and expanding operational capability.

From first daily driverless operations to broader conditions

Urmson said Aurora’s focus is to deliver self-driving technology “safely, quickly, broadly” with trucking as the first market. He said that in 2025 Aurora launched what he described as the first driverless trucks operating daily in the U.S., adding that Aurora is “the only company operating them regularly every day.” Over the course of the year, he said the company expanded from daylight-only operations on a particular route in good weather to operating “day and night across a variety of routes” including in rain and higher winds.

One operational lesson, he added, was that adverse weather constrained early utilization more than expected. Urmson said it “rains a lot in Texas” and that around 40% of days in the past year the company was not able to operate due to various weather conditions. He said Aurora now has the ability to operate in rain and expects to continue improving the level of rain and environmental conditions it can handle through 2026, describing this as important for serving customer reliability.

2026 scale-up plans: hundreds of trucks and new hardware

Urmson outlined what he framed as a step change in deployment. Aurora currently operates “a handful” of driverless trucks daily, he said, but expects to be running “a couple of hundred trucks” driverlessly by the end of 2026. He also said the company plans to launch its second-generation hardware in Q2 on a new truck platform from International, operating driverlessly “nobody on board” while serving customers.

He said Aurora expects to expand across the Southern U.S. during the year and sees mapping as an operational task rather than a technical barrier. On scaling routes, he said Aurora has “literally no concerns” about its ability to scale maps and cited the company’s pace of expanding lanes from one to four to ten.

On vehicle volume scaling, Urmson pointed to an upfit strategy, noting a partnership with ROUSH to upfit International trucks with a by-wire overlay and Aurora’s second-generation hardware kit. He said Aurora expects, in the back half of 2026, to be producing 20 trucks per week for deployment.

Revenue expectations and customer demand

Urmson said Aurora’s couple hundred driverless trucks are expected to support an $80 million revenue run rate by the end of 2026, with revenue driven by miles and utilization. He described Aurora’s current model as owning and operating trucks and “effectively” operating like a trucking company, but “under the flag of our partners.”

He also highlighted what he described as rising customer interest as Aurora moved from demonstration to real driverless operations. Urmson said discussions have shifted from hypothetical economics to practical deployment, and that Aurora has “more demand than we can possibly supply” in the near term. He added that some customers have asked for exclusivity, which he said Aurora does not provide, though the company aims to reward long-time partners.

As an example of customer pull, Urmson discussed a partnership with Detmar. He said Detmar approached Aurora after seeing progress and press coverage, seeking to move sand between a mine and distribution center using freeway routes. He characterized the use case as a “virtual conveyor belt,” where greater truck utilization translates directly into more material moved.

Cost, hardware roadmap, and capital-light manufacturing model

Urmson said Aurora views hardware supply chain and cost reductions as competitive advantages as the company progresses through second- and third-generation systems. He said Aurora’s second-generation hardware will be produced by Fabrinet, supporting up to about 1,500 units. He added that moving from the current hardware to gen two is expected to cut component costs in half and triple durability, and that the later move to a third-generation system produced with partner Imovio is expected to reduce cost by another factor of two.

Urmson described the Imovio arrangement as an incentive-aligned structure in which Imovio finances the hardware and Aurora pays based on utilization. He said Imovio is investing roughly $350 million in engineering and manufacturing capability, and characterized the model as avoiding a “CapEx cliff” for Aurora.

When asked about margin and cost visibility, Urmson said Aurora has strong visibility into gen-two costs because it is manufacturing today and has supply agreements in place. He said he is “more confident” in the gen-three cost outlook given Imovio’s manufacturing and supply chain capabilities, noting ongoing work reviewing bill of materials and assembly costs.

Technology approach: maps, multimodal sensing, and “Verifiable AI”

Urmson emphasized Aurora’s “Verifiable AI” approach, which he said is critical for safety and building conviction in performance. He also addressed industry debates about the use of HD maps and sensor configurations.

  • HD maps: Urmson argued computers are strong at storing and recalling information and said maps help a vehicle anticipate what is coming rather than only reacting in real time. He said Aurora’s map-building is now “not a technology problem,” but an operational process the company can scale.
  • Sensor suite: Urmson said Aurora believes in multimodal sensing—cameras, radar, and lidar—because each modality has different failure modes and strengths. He cited benefits in night driving and adverse conditions such as dust and weather.
  • Simulation vs. real-world testing: He said simulation is important but must be grounded in reality and properly evaluated, including accurate sensor statistics such as lidar return probabilities and intensities. He said Aurora prioritizes data quality, targets real-world collections where needed, and uses simulation to safely create scenarios that would be difficult to generate in the real world.

On regulation, Urmson said Aurora does not see regulations as a barrier to building the business today, noting that driverless operations are permitted in most states, including across the southern corridor. He also said Aurora sees supportive signals at both state and federal levels and referenced positive statements about automated vehicles from current federal leadership.

Urmson said Aurora ended the year with $1.5 billion in cash and believes it has sufficient liquidity to reach free cash flow on a run-rate basis, while noting the company may use its ATM program to fund tax obligations on RSUs and employee bonuses. Looking ahead, he said he is most excited about putting second-generation hardware and a new truck platform on the road at scale, suggesting that by the end of 2026 road-trippers in the Southern U.S. “can’t help but see” Aurora’s trucks operating driverlessly.

About Aurora Innovation (NASDAQ:AUR)

Aurora Innovation, Inc is a technology company specializing in the development of self-driving vehicle systems for both passenger and commercial applications. Headquartered in Mountain View, California, Aurora has built an end-to-end platform—known as the Aurora Driver—that integrates proprietary software, machine learning algorithms and a suite of sensors (LiDAR, radar and cameras) to enable vehicles to operate safely and efficiently in diverse driving environments.

The company’s core business revolves around designing, testing and deploying its autonomy stack on vehicles from established automotive and transportation partners.

Featured Articles