The transformation of autonomous trucks is more likely to come “a bit later,” but may arrive earlier than the market expects

Overview

This report discusses the potential impact of AI, especially physical artificial intelligence, on autonomous trucks and global truck OEMs. It argues that the truck industry has historically adopted technology about 10 to 20 years more slowly than passenger cars, but the commercial logic for autonomous trucks is very clear: driver shortages, driver hour restrictions, low asset utilization, fuel costs, and insurance costs all provide economic justification for automation. Bernstein’s core judgment is that autonomous trucks will not change the industry overnight, but the question is more likely one of “when” rather than “whether.”

Core views

The report’s core view is that autonomous trucks offer real and quantifiable cost savings for freight operators, but their impact on traditional truck OEMs is more complex. In the short term, OEMs still retain advantages in chassis, dealerships, maintenance, and customer relationships, and may also benefit from higher vehicle prices, recurring revenue, and potential incremental equipment demand; in the medium to long term, large fleets may be the first to adopt autonomous driving thanks to TCO advantages and gain share, leading to consolidation in the freight industry and weakening OEM profit and pricing advantages with small customers. The report therefore remains relatively cautious on European truck OEMs.

Analysis framework

The report uses its proprietary truck TCO model to break down the costs of human-driven Class 8 trucks versus autonomous trucks, focusing on driver costs, fuel, maintenance, vehicle depreciation, and additional operating costs such as sensors and redundant hardware systems, data and cybersecurity, shuttling, and inspections. The report also tracks the funding cycle of autonomous trucking companies, changes in AI compute and model architectures, OEM partnerships, regulatory progress, and the competitive landscape of key players.

Methodology notes

• 成本模型Total Cost of Ownership

  TCO

  The report uses total cost of ownership per mile as the core metric, comparing driver, fuel, maintenance, depreciation, and additional autonomous-driving operating costs between human-driven trucks and autonomous trucks.

• 技术路径End-to-End Autonomous Driving

  E2E autonomous driving

  The report believes advances in AI are pushing the industry from rule-based models toward end-to-end models, improving scalability and reducing reliance on expensive sensor hardware.

• 技术路径Vision-Language-Action

  VLA

  The report views VLA as the next-stage direction for autonomous-driving models, combining large language model capabilities for text and video understanding, which may improve scenario generalization.

• 行业结构Customer Base Consolidation

  freight customer consolidation

  The report believes the TCO advantage of autonomous driving may allow large fleets to adopt first and expand market share, thereby changing OEM customer structure and profit structure.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Daimler Truck AG (DTG.GR)

  Traditional truck OEM, while also developing autonomous driving capabilities in-house through Torc Robotics

  Strengths

  Has OEM chassis, manufacturing, dealer, and maintenance systems, and is one of the few traditional OEMs developing virtual driver models in-house.

  Weaknesses

  Autonomous-driving investment has already led to substantial losses, and the report rates it Underperform.

  Comparison

  Compared with other OEMs, Daimler Truck is more vertically integrated on the in-house development path, but its cash burn and execution risks are also higher.

  Risks

  Customer consolidation, rising R&D spending, delays in autonomous-driving commercialization, and valuations not reflecting disruption risk.

• Traton SE (8TRA.GR)

  European truck OEM, affected by autonomous-driving and electrification trends

  Strengths

  Can benefit from its existing commercial vehicle manufacturing and customer service network.

  Weaknesses

  The report believes European OEMs as a whole face greater pressure from higher investment needs and customer structure changes.

  Comparison

  Rated Market-Perform, more neutral than Daimler Truck.

  Risks

  Greater bargaining power of large fleets, industry consolidation, and rising technology investment requirements.

• Volvo AB (VOLVB.SS)

  European truck OEM, positioned within the chain of potential autonomous-truck impacts

  Strengths

  Has a strong commercial vehicle brand, engineering capabilities, and maintenance network.

  Weaknesses

  The report believes the share price and valuation have not yet fully reflected the challenges of the next cycle.

  Comparison

  Rated Market-Perform, with a target price of SEK 290.

  Risks

  Autonomous driving driving customer concentration, electrification capex, and pressure on aftermarket service profits.

• Paccar Inc (PCAR)

  U.S. truck OEM, listed by the report as a relatively preferred stock

  Strengths

  Has strong U.S. market exposure and existing OEM ecosystem advantages; the report rates it Outperform.

  Weaknesses

  Still faces disruption to traditional profit pools from autonomous driving and electrification.

  Comparison

  More favored among the U.S. companies covered in the report, with a better rating than Cummins.

  Risks

  If autonomous-driving platforms and large fleets capture more of the value distribution, OEM margins may come under pressure.

• Cummins Inc (CMI)

  Powertrain and related commercial vehicle supply chain company

  Strengths

  Holds an important position in commercial vehicle power systems and components.

  Weaknesses

  The report rates it Market-Perform, with a target price below the current price.

  Comparison

  Compared with Paccar, the report is more neutral on Cummins.

  Risks

  Electrification, changes in autonomous-driving architectures, and OEM capex cycles may affect demand and valuations.

• Aurora Innovation

  Autonomous trucking technology company and representative of virtual driver services

  Strengths

  Already testing on the road with customers, with a long-term commercialization timeline pointing to 2027, and has presented a relatively aggressive DaaS revenue vision.

  Weaknesses

  Its per-mile revenue assumptions are more aggressive than peers, and commercialization and unit economics still need to be proven.

  Comparison

  It accounted for a high share of early-stage funding, but new funding since 2025 has been spread more broadly across multiple competitors.

  Risks

  Excessively high software fee pricing could affect fleet adoption, while regulation, reliability, and scaling remain key risks.

• PlusAI; Waabi; Kodiak Robotics; Bot Auto; Einride

  Autonomous trucking competitors and beneficiaries of the AI capital cycle

  Strengths

  Benefit from AI models, the NVIDIA ecosystem, participation by industrial capital such as Uber, and a new funding cycle.

  Weaknesses

  Most companies are still in the early stages of commercialization and need to prove safety, scalability, and profitability models.

  Comparison

  The report notes that funding since 2025 has been more diversified than in the previous cycle and is no longer highly dominated by Aurora.

  Risks

  Commercialization delays, reversal in capital market sentiment, uncertainty in OEM partnerships, and regional regulatory restrictions.

Key data

• Driver cost share>40% TCODrivers are the largest cost item for truck operators, at about $1.0 per mile.
• Autonomous driving savings potential$0.65/mile, about 28% TCO reductionThis is the theoretical saving before paying virtual driver software fees.
• Virtual driver software fee assumptionabout $0.3/mileThe report believes that in a competitive market, pricing needs to stay below the theoretical savings ceiling so that fleets can retain part of the benefit.
• Fleet TCO improvement after adoptionabout 15%This may still be achieved after paying virtual driver software fees and other incremental autonomous-driving operating costs.
• Historical autonomous trucking fundingabout $7.5bn-$7.6bnTotal historical equity funding for the basket of autonomous trucking companies tracked by the report.
• New funding since 2025about $2.6bnAbout one-third of the historical funding of the relevant companies, reflecting a new AI-driven capital cycle.
• U.S. driver shortageabout 80,000 in 2021; about 160,000 in 2030The chart shows that the shortage in 2030 could be about twice the 2021 level.
• Long-haul mileage structure>600 miles accounts for more than 60% of expected mileageLong-haul highway scenarios are better suited for early autonomous-driving commercialization.
• Aurora caseRevenue per truck per week rises from $6,150 to $16,400; profit rises from $185 to $1,695From an Aurora example, showing autonomous driving can improve margins through higher utilization and lower costs.
• NVIDIA compute improvementThor platform compute is about 4x that of Orin-XStronger edge compute helps in-vehicle perception and decision-making while potentially reducing chip count.

Impact & implications

From an investment perspective, the autonomous trucking theme is a more direct positive for software and autonomous-driving platform companies, but not a linear positive for traditional OEMs. OEMs still need to exist and provide redundant chassis, maintenance networks, and integration capabilities, but as industry customers concentrate from many small fleets toward large fleets, OEM pricing power and aftermarket profit pools may be compressed. The report believes current truck stock valuations do not clearly price in the risk of structural disruption, and the next cycle may be more difficult than the last because electrification and autonomous driving together raise R&D and capital expenditures while bringing more competition.

Risks

• The commercialization timeline for autonomous trucks may continue to shift later, putting pressure on the capital cycle and valuations.
• Regulatory approvals are currently concentrated mainly in Texas and the Sunbelt in the U.S., while Europe remains several years behind.
• If virtual driver software fees exceed what fleets can accept, TCO improvement may be weakened and adoption slowed.
• If large fleets are the first to adopt autonomous driving, freight industry concentration may rise, weakening OEM profit advantages with small customers.
• Autonomous driving may reduce accidents and repair demand, potentially eroding the high-value aftermarket business of traditional OEMs.
• The simultaneous push toward electrification and autonomous driving will raise OEM R&D and capital expenditures, suppressing cash generation.
• Additional costs from sensors, redundant systems, cybersecurity, remote support, and shuttle operations may be higher than expected.

What to watch

• Whether Aurora, PlusAI, Torc, Waabi, and others can advance commercialization as planned around 2027.
• Progress in regulatory approvals for autonomous trucks outside Texas and the Sunbelt in the U.S.
• Whether real-world virtual driver software pricing is closer to $0.3 per mile or to Aurora’s more aggressive assumption of $0.85 per mile.
• The speed of adoption by large fleets and its pressure on the market share of smaller freight operators.
• Real-world performance of NVIDIA Thor, Blackwell, and VLA foundation models in in-vehicle compute, cost, and generalization ability.
• Partnerships between traditional OEMs and autonomous-driving software companies, as well as whether losses narrow along the Daimler Truck/Torc in-house development path.
• Whether truck stock valuations begin to reflect the structural risks brought by autonomous driving, electrification, and customer consolidation.