Atlas Energy Solutions hauls over 450,000 tons of sand with Kodiak driverless trucks

Atlas Energy Solutions supplies proppant, which is critical to well completions, to oil and gas producers and hauls sand to well sites across the Permian Basin. Its stated approach is to find critical energy infrastructure with inefficiencies and engineer solutions that improve efficiency, reduce risk and enhance customers' operational success. Until now its driverless trucks operated from one load-out point at a time, while customer demand was growing.

Atlas works with Kodiak AI on a driverless proppant delivery program, using trucks equipped with the AI-enabled Kodiak Driver, Kodiak's autonomous driving system; initial driverless deployments began in 2024. Atlas has added a second simultaneous load-out point along its 42-mile Dune Express sand conveyor system, so driverless trucks now operate from load-out sites in Texas and New Mexico about 90 minutes apart. Earlier this year, Atlas and Kodiak also introduced capabilities that let one autonomous truck haul three connected trailers at once.

As of March 31, 2026, Atlas operated 28 driverless trucks across 15 distinct routes, which together have carried approximately 7,000 loads and hauled more than 450,000 tons of sand, and the fleet completed over 23,500 driverless hours in the first quarter of 2026. On July 20, Kodiak-powered trucks delivered 176 loads of sand in a single day, an all-time daily high across 18 months of driverless deployments with Atlas. Atlas and Kodiak have agreed a timetable to grow the fleet to 100 autonomous trucks by mid-2027 and expect public-road operation by early 2027, subject to regulatory and operational milestones.

• Adding a second load-out point let the driverless fleet serve a wider area at the same time.
• The program has scaled steadily since its first driverless deployments, a pattern of staged growth rather than a single large rollout.
• Plans for a larger fleet and public-road operation remain subject to regulatory and operational milestones, so treat them as targets rather than results.