Tech & AI Oct 19, 2026 · 7 min read

Daimler puts a date and a price model on driverless trucks

Daimler says the driver leaves its first Torc truck by end-2026. Daimler Truck CFO Eva Scherer said the company expects to take the driver out on its first route by the end of this year, then scale toward 2028, FreightWaves reported on 17 October. The truck is sold or financed as usual. The Torc “virtual driver” is billed per mile as software. Our view: the per-mile price matters more than the date. One route is not capacity. A per-mile fee is a number that will end up in linehaul quotes. Our call: no public per-mile price from Daimler or Torc before mid-2027, and no driverless effect on Texas spot rates in 2027.

Daimler’s driverless date is one route; its per-mile price is the part that will reach your quotes
Daimler Truck expects to take the driver out of its first autonomous truck route by the end of 2026. That is about ten weeks from now. CFO Eva Scherer gave the timeline at a session at Daimler Truck Financial Services’ new headquarters in Charlotte, North Carolina, FreightWaves reported on Saturday 17 October. In FreightWaves’ words, Scherer said the company “expects to get the driver out on its first route by the end of this year, then build mileage from there, targeting commercialization at scale toward 2028.”
Our view: the date is the least important part. A single driverless route will move a tiny share of US freight. The important part is how Daimler plans to charge for it. The truck is sold the normal way. The software that drives it is billed by the mile. That is a pricing unit, and pricing units are what eventually show up in a forwarder’s linehaul quotes. Linehaul means the long-distance leg between cities or terminals.

What Daimler actually said

FreightWaves describes the system in plain terms. The platform is “an autonomy-ready Freightliner Cascadia with full redundancy in braking and steering.” Redundancy means a backup system takes over if the main one fails. That matters because no driver is there to grab the wheel. The Cascadia is paired with “a virtual driver from Torc Robotics, the company’s autonomous subsidiary.” A virtual driver is the software and sensors that do the driving.
The business model is the part to note. Customers buy or finance the truck as they do today. Then, in FreightWaves’ words, “the virtual driver comes as a per-mile software-as-a-service layer on top.” Software-as-a-service, or SaaS, means you pay for use instead of owning the software outright. Daimler also said it “will finance whatever shape the autonomous model ultimately takes.” In other words, the company’s lending arm plans to fund these trucks however carriers end up buying them.
The report says the system is modular, not locked to one autonomy provider. It also says Scherer described “a shift from older imitation-based training to reinforcement learning”. Imitation training teaches software to copy human drivers. Reinforcement learning lets it learn by trial and reward, so trucks can cope with routes they have not seen. Even so, FreightWaves wrote that autonomy “will still roll out on defined lanes that expand gradually, with safety as the gating factor.”
The article includes no direct quote from any Daimler executive on autonomy. Every statement above is FreightWaves’ summary of what was said.

Torc has been heading for Laredo for some time

None of this comes out of nowhere. Torc has been building toward a Texas launch for several years. Daimler’s 17 October comments fit the timeline Torc itself has given.
In an interview published in July 2024, Torc’s vice president of commercialization, Nick Elder, told Trucking Dive the company aimed to take long-haul drivers off public roads in Texas in late 2026. He said early deployments would be narrow: “When we talk about initial deployments, we’re going to be looking at one or two or three lanes.” Torc had said in August 2023 that it expected to be ready for a product launch in 2026, “with scalable market entry in 2027.”
In January 2025, FreightWaves reported that Torc was winding down its Albuquerque test site and its Stuttgart tech hub. Instead it would focus on Dallas-Fort Worth. Chief commercial officer Andrew Culhane said the company was shifting its focus to testing on its launch lane from DFW to Laredo. The company’s release added that “Laredo is the largest economic port of entry in the US, with more than 15,000 truck crossings per day.”
In May 2025, Torc opened its first autonomous hub on an 18-acre site at AllianceTexas in Fort Worth, Trucking Dive reported. At the time, the company was talking about a “2027 U.S. market launch.”
So the story holds together. Driver out on one lane late in 2026. Commercial scale in 2027 and 2028. Daimler’s CFO did not name the first route, according to the FreightWaves report. Torc’s own named launch lane is I-35 from Dallas-Fort Worth to Laredo.

Daimler is the third name on the 2026 list

Daimler is not first. It is joining two rivals that have already set 2026 targets.
Aurora is furthest along. In its second-quarter results on 29 July 2026, Aurora said it was “fully allocated to exit the year with 200 driverless trucks in operation.” CEO Chris Urmson said the company was “positioned to put hundreds of driverless trucks on the road this year.” That release cited “hundreds of thousands of driverless miles” but no exact total. Earlier, in February, Breakbulk News reported that Aurora had passed 250,000 fully driverless miles. It also reported that Aurora projected 2026 revenue of $14 million to $16 million, and expected revenue to exceed expenses from 2028.
Kodiak is going after a different lane. Trucking Dive reported on 29 September that Kodiak aims to start unsupervised long-haul service by the end of 2026. That means no safety observer in the cab. The lane is the 219-mile I-45 run between Dallas-Fort Worth and Houston, with Ikea as launch customer. Kodiak says it has delivered more than 1,300 loads for Ikea and logged more than 750,000 autonomous miles for the retailer with an observer on board. Its industrial fleet in the Permian Basin, run by Atlas Energy Solutions, has already reached the readiness level Kodiak sets for driverless service. CEO Don Burnette said in a 25 September release: “Our progress since beginning our final push towards driverless commercial deliveries has been incredible.”
Put the three side by side and the pattern is clear. Every company starts in Texas, on a few named lanes, with a handful of customers. Aurora’s 200-truck year-end goal is the largest stated number. Even that is small next to the US trucking fleet.

Why the per-mile model matters to forwarders

Here is our reading of what this means for people who buy trucking.
First, the cost of a driverless load will be made of familiar parts. There is the truck, financed as usual. There is fuel. And there is a software fee charged per mile. That is easier to model than a driver’s pay, which moves with the labour market, hours-of-service rules and turnover. When carriers start offering driverless capacity, expect it to come as a firm per-mile price on a fixed lane. Spot-market swings will play a smaller part.
Second, the lanes are narrow and they are all in Texas. If you move cross-border freight through Laredo, the I-35 corridor to Dallas-Fort Worth is the lane to watch. If you move import boxes off the Port of Houston, I-45 to Dallas is Kodiak’s lane. Freight elsewhere will see nothing from this in the next year.
Third, the first users will be carriers, not shippers. Daimler sells trucks to fleets. Torc’s software rides on those trucks. A forwarder will meet driverless capacity through a carrier or a broker that runs it, not by signing a software deal itself. Ask your Texas carriers now whether they are in any of these pilots. The answer tells you who will price these lanes first.
Fourth, do not plan around it yet. Daimler’s reported target puts scale “toward 2028”. Torc’s earlier plans said one to three lanes at first. Aurora is guiding to 200 trucks. Kodiak is still in final checks: Trucking Dive reported its readiness score for long-haul at 93% against a 100% threshold. None of this changes truck capacity on most US lanes in 2027.

What to watch, and our call

Three signals will tell you whether this is speeding up.
One is whether Daimler or Torc actually announces a driver-out run before 31 December, and on which lane. Two is whether a per-mile software price is ever made public. So far it has not been. Three is whether Aurora hits its 200-truck year-end figure, which it is likely to report with its fourth-quarter results.
Our prediction: Torc will run a truck with no one in the cab on one Texas lane by the end of the first quarter of 2027, most likely I-35 between Dallas-Fort Worth and Laredo. It may slip a few weeks past Daimler’s year-end target. We expect no public per-mile price from Daimler or Torc before mid-2027. And we expect no visible driverless effect on Texas truckload spot rates through the end of 2027. The first real price signal for forwarders will come from carriers quoting fixed per-mile rates on these named Texas lanes. We expect the first such quotes in 2028, in line with Daimler’s own scale target.

Sources