Kodiak AI Driverless Trucks: DTL Deal Explained ⚡
Learn how Kodiak AI driverless trucks and DTL Transport enable Driver-as-a-Service—breaking down revenue logic, risks, and milestones. Get the DTL deal details.
Kodiak's 75-Truck Deal Tests a New Model for Autonomous Freight
Autonomous trucking companies have spent years trying to prove they can move freight with driverless technology. The harder part is making it a business customers will actually pay for - something they can plug into their day-to-day operations, and that can grow without the tech provider turning into a full-on trucking operator.
Kodiak AI's agreement to supply 75 driverless trucks to DTL Transport is an important step toward that kind of repeatable model. DTL expects deliveries to start in 2028 and plans to use the trucks on interstate routes. This deal also differs from Kodiak's current Trucking-as-a-Service approach: under the new arrangement, the customer owns and runs the trucks and licenses Kodiak's technology.
That difference changes the economics and the responsibility split - who pays for the vehicles, how Kodiak gets paid, and what has to go right for the model to scale. Still, it doesn't, by itself, prove that driverless highway trucking is ready for broad deployment. The contract value hasn't been disclosed, and the trucks won't arrive for years.
Here's what the agreement changes, how Kodiak's potential revenue has been estimated, and which operational milestones matter more than the headline number of trucks.
What did Kodiak and DTL actually agree to?
Kodiak will deliver 75 self-driving trucks to DTL Transport, with deliveries expected to begin in 2028. The companies announced the binding commercial agreement on October 28, 2026. Kodiak says this is its first binding deal for driverless trucks intended to operate on public highways.
DTL is a California-based mid-market carrier that hauls fresh produce and other goods to grocery chains. It plans to deploy the trucks across multiple interstate routes, including Dallas to El Paso and Interstate 20 between Dallas and Atlanta. Kodiak has already tested and operated on both routes, so the agreement has a clearer operating context than a contract tied to unknown future lanes.
The companies also completed a pilot in California the previous month. During that trial, a Kodiak-equipped truck hauled perishable goods for DTL for about 225 miles between Fresno and Los Angeles, using Interstate 5 and State Route 99. A human safety operator remained behind the wheel. In other words, the pilot showed a working freight operation with a safety driver - it didn't show that the same kind of route work had already been done without one.
That matters because a pilot and a driverless commercial service are different steps. A pilot is about testing how the technology behaves on a route and whether it fits into a customer's logistics process. A driverless commercial operation also has to meet the carrier's safety and operational requirements without relying on a person in the cab as a fallback.
Kodiak's next milestone is to remove human safety operators from the cab and start a driverless highway service in December, carrying goods for Ikea on Interstate 45 between Dallas and Houston. That service is separate from the DTL agreement. DTL's trucks aren't expected to start arriving until 2028, so the near-term launch and the later supply deal shouldn't be treated as the same event.
For now, the deal lays out three clear points: a carrier has committed to a defined number of trucks, the intended use includes public-highway routes, and the customer relationship has moved beyond a trial. The agreement doesn't disclose the contract price, delivery timing for each vehicle, or detailed terms covering service, maintenance, and performance - those details will determine the real economics once deployment begins.
Why does the deal change Kodiak's business model?
Kodiak currently earns revenue through Trucking-as-a-Service (TaaS). In that setup, Kodiak owns and operates autonomous trucks as a third-party carrier for customers. It moves freight and collects revenue from daily operations, while the customer pays for transportation rather than buying the self-driving system itself.
Kodiak lists TaaS customers including Charger USA, Roehl Transport, and Werner Enterprises. Its ongoing freight operations for those customers run on three active Texas routes, but founder and CEO Don Burnette has described them as still being in the pilot stage. With TaaS, Kodiak has direct control over how the trucks are deployed, which can help it prove the technology and learn how the system behaves in live operations.
That control comes with responsibilities. If Kodiak owns and runs the trucks, it also takes on more of the trucking operation - not just the software. The business has to handle the real work of operating a carrier: coordinating freight, managing vehicles, and building the capacity needed to serve customers. Kodiak can earn transportation revenue, but it also carries more operational workload and capital requirements.
The DTL agreement points toward Driver-as-a-Service (DaaS). In that model, the customer owns and operates the trucks, while Kodiak licenses its autonomous-driving technology and charges fees based on miles driven over the contract term. Kodiak's role shifts toward supplying and supporting the technical system, rather than acting as the freight carrier.
Kodiak has said it wants to head in this "asset-light" direction. The basic trade-off is straightforward: an asset-light supplier may avoid owning a large fleet, but it still needs to show that its technology works reliably in a customer's environment - and that the customer can operate it effectively. Kodiak gives up some control over fleet operations in exchange for a model it hopes will be easier to expand across customer-owned trucks.
Moving from TaaS to DaaS isn't just about how invoices are structured. TaaS can work as a proving ground: Kodiak controls the truck and operating process, gathers experience, and shows prospective customers how the system works. In DaaS, the customer takes on more of the day-to-day operating responsibility. That transition only works if the technology is mature enough and the customer has the capacity to integrate it into its fleet.
Kodiak already has a DaaS customer in Atlas Energy Solutions, where the trucks operate off-road in West Texas's Permian Basin. Atlas has owned and operated 35 driverless trucks equipped with Kodiak's system since 2024, and Kodiak says it expects the fleet to grow to as many as 100 by mid-2027. The DTL deal is different because it brings the same model to on-highway freight - public interstates and more conventional logistics routes.
What could the 75 trucks mean for revenue?
Kodiak hasn't disclosed the value of the DTL contract. So any revenue estimate here is an illustration based on assumptions - not a confirmed figure for what Kodiak will collect.
In a 2025 investor presentation, Kodiak forecast annual recurring revenue of $150,000 to $190,000 per truck. The assumptions included a licensing fee of about $0.97 per mile and annual mileage of roughly 225,000 miles per truck. The presentation also relied on other operating assumptions, so the per-truck number should be treated as a projection rather than a guaranteed rate.
If you apply the low and high ends of that forecast to 75 trucks:
- 75 trucks × $150,000 in annual recurring revenue per truck = $11.25 million per year.
- 75 trucks × $190,000 = $14.25 million per year.
That math is why the deal could work out to at least about $11.2 million in annual recurring revenue once all 75 trucks are delivered and operating under similar assumptions. It doesn't mean that amount starts immediately, that all 75 generate revenue right away in 2028, or that Kodiak is guaranteed to earn it for the full life of the agreement.
There are dependencies between a signed contract and realized revenue. The trucks have to be delivered and placed into service. They then need to run enough miles to support the mileage-based revenue model. Kodiak and DTL also have to get through integration and operational details, and DTL's deployment has to match the assumptions used in Kodiak's forecast.
The start date is part of the issue, too. Deliveries are expected to begin in 2028, but the information available doesn't say how quickly all 75 vehicles will arrive or when each one will start generating recurring revenue. A phased rollout would mean revenue ramps over time rather than looking like a full 75-truck run rate on day one.
The agreement also matters as a reference point for the market. A mid-market carrier committing to a defined highway fleet is more concrete than a pilot alone. But one customer agreement still doesn't tell you how big the broader demand is, and it doesn't show what margins Kodiak can earn when the trucks are actually running.
Kodiak shares have fallen 87% from their opening-day price, which keeps investor attention tightly focused on execution. This deal could become a benchmark for how well Kodiak moves from pilots to customer-owned deployments. The more telling metrics will be delivery progress, active trucks, miles driven, and whether the projected per-truck economics hold - not the truck count alone.
What must happen before autonomous trucking scales?
The agreement makes the customer relationship more concrete, but the real work is what happens between signing and sustained operation. Driverless trucking has to fit into real freight schedules, cover the routes customers need, and deliver enough reliability to justify changing how fleets staff and manage trucks.
Safety and driverless operation
The California pilot with DTL used a human safety operator. Kodiak's planned December driverless service for Ikea is separate, and it's the milestone intended to test operations without a person in the cab. If that launch goes well, it would be relevant evidence - but it still wouldn't automatically validate every route or every deployment.
DTL's trucks are intended for different routes, and their delivery isn't expected until 2028.
For a buyer, the question isn't only whether the system can drive. It's whether the overall operation meets the customer's safety process and can handle the conditions on its routes. The agreement makes that question more actionable, but the available details don't provide performance figures or specific operating thresholds.
Fleet integration and operational ownership
Under DaaS, DTL owns and operates the trucks. That means the technology has to fit into the carrier's actual fleet operations, rather than remaining something Kodiak manages as a demo. Things like route planning, dispatch, vehicle availability, and how responsibilities are divided between carrier and technology provider all affect whether the service works in practice.
Burnette has argued that mid-market carriers can move quickly and integrate new technology more easily than larger operators. That could make DTL a practical early customer. But it doesn't prove that every carrier will find onboarding equally easy. A deployment that depends on close involvement from Kodiak's team could work with one partner and be harder to reproduce across many fleets.
Utilization and unit economics
Kodiak's revenue assumptions depend on trucks accumulating substantial mileage - about 225,000 miles per year in the company's forecast. If a truck is delivered but not used much, it won't support the same mileage-based revenue model. Utilization also determines how much value the customer gets from buying and operating the vehicle.
That's why operational data matters more than the idea of a fleet size. Investors and other potential customers will want to see how many trucks are active, how many miles they run, how consistently they're available, and whether the economics match the original assumptions. Those results aren't available for DTL yet, since the trucks haven't been delivered.
Repeatability beyond one customer
A technology supplier can only scale if it can support deployments beyond a single customer without rebuilding the operating process from scratch every time. The DTL deal tests whether Kodiak can move from a company-operated proof point to a customer-operated highway fleet.
Atlas provides an example of a customer-owned model off-road, and DTL extends that idea to public interstates. The routes Kodiak has already tested give a starting point, but repeatability will require more than familiar segments of road. Each customer has its own fleet, workflows, and freight requirements. If each deployment needs extensive custom work, the asset-light model may still involve heavy support costs. If integration becomes predictable, the licensing approach has a clearer path to scale.
Key takeaways
- The DTL agreement is for 75 trucks, with deliveries expected to start in 2028; it isn't an immediate 75-truck deployment.
- The contract is set up as Driver-as-a-Service, shifting truck ownership and day-to-day operation to the customer.
- Kodiak's revenue estimate depends on mileage and other assumptions, and the contract's value hasn't been disclosed.
- The most important proof will be safe operations, customer integration, utilization, and repeatable deployments - not the announcement alone.
Conclusion
Kodiak's agreement with DTL moves its public-highway plans toward a customer-owned operating model. Instead of providing transportation with trucks that Kodiak owns, Kodiak aims to license its technology to a carrier that will own and run the vehicles.
That could make the business more asset-light - but the result will depend on execution. The revenue estimate relies on high annual mileage, the trucks aren't expected until 2028, and the contract value remains undisclosed. In the meantime, the main question is whether driverless service works reliably, whether DTL can integrate the system into its fleet, and whether the model can be repeated without a lot of extra support.
If you're trying to judge the deal, the safest way to look at it is as a credible commercial milestone, not proof that autonomous trucking has already scaled. Track what happens on the ground - operating trucks and miles driven - not just signed orders. Those are the signals that will show whether Kodiak's technology can become a durable supplier business.