# Mythos AI assists towboat crews as Aurora deploys fully driverless trucks

> Source: <https://runtimewire.com/article/mythos-ai-aurora-us-freight-automation>
> Published: 2026-08-14 08:11:57+00:00

Geoff Douglass's Mythos AI entered commercial freight with software that advises a Mississippi River towboat crew while leaving the vessel under human control. Chris Urmson's [Aurora Innovation](https://ir.aurora.tech/company-information/leadership-team?ref=runtimewire) is deploying Class 8 trucks in Texas without anyone behind the wheel. The distinction shapes what each founder must prove: Mythos needs measured gains from a crew-assistance installation, while Aurora needs revenue and reliable utilization from fully driverless operations.

The two approaches feature in [an August 14 Reuters analysis](https://www.reuters.com/commentary/reuters-open-interest/smart-tugs-trucks-that-dont-sleep-automation-rewires-us-transport-2026-08-14/?ref=runtimewire) of automation across US freight. Reuters reported that Southern Devall installed Mythos AI's Advanced Pilot Assist System, or APAS, on a commercial Mississippi River towboat in 2025. [The Waterways Journal attributed its account of the installation to a Southern Devall statement](https://www.waterwaysjournal.net/2025/08/08/southern-devall-installs-mythos-ai-system-on-vessel/?ref=runtimewire). Aurora began commercial freight operations in April 2025 and expanded driverless operations across Sun Belt routes in 2026. The companies' disclosures show two different routes into freight autonomy.

### Douglass starts with the captain

A [CoMotion profile](https://www.comotionglobal.com/speakers/geoff-douglass?ref=runtimewire) says Douglass spent 17 years developing unmanned maritime craft. [Andreessen Horowitz's American Dynamism 50](https://a16z.com/american-dynamism-50-ai/?ref=runtimewire) identifies Douglass, John Houston and Allen Flick as the founders of West Palm Beach-based Mythos AI and says they started the company in 2020. A [2022 University of Memphis conference biography](https://www.memphis.edu/setwc/2022speakerbios.pdf?ref=runtimewire) says Douglass had commissioned more than 40 unmanned maritime systems before founding Mythos.

Mythos AI's APAS monitors hazards, tracks vessels, calculates stopping distances, analyzes river conditions and identifies fuel-saving strategies, [according to Reuters](https://www.reuters.com/commentary/reuters-open-interest/smart-tugs-trucks-that-dont-sleep-automation-rewires-us-transport-2026-08-14/?ref=runtimewire). [The Waterways Journal reported](https://www.waterwaysjournal.net/2025/08/08/southern-devall-installs-mythos-ai-system-on-vessel/?ref=runtimewire) that the installation gives the captain information and alarms while leaving control of the towboat with the crew.

Southern Devall provides a sizable test fleet. [The company says](https://www.southerndevall.com/About/?ref=runtimewire) it operates more than 70 towboats and 230 barges with about 700 mariners. The inland-waterway towing company is headquartered in Sulphur, Louisiana, and operates across Louisiana, Texas, Tennessee and other locations. Mythos AI and Southern Devall have not published independently measured fuel savings or safety improvements from the APAS installation.

The potential fuel benefit starts with the efficiency of river transport. A US Army Corps of Engineers comparison cited by Reuters found that one gallon of fuel can move one ton of cargo more than 500 miles by barge, compared with about 60 miles by truck. Better stopping calculations, propulsion decisions and traffic planning could reduce waste, although the amount attributable to Mythos AI remains unquantified.

Douglass is also taking Mythos AI beyond advisory software. In September 2025, Mythos AI [said its fully autonomous MNAV system](https://www.prnewswire.com/news-releases/mythos-ai-surpasses-11-000-miles-logged-in-full-autonomy-a-world-first-milestone-in-shipping-navigation-302563810.html?ref=runtimewire) had logged more than 11,000 miles and completed over 15,000 encounters compliant with international collision-avoidance rules. Those figures were supplied by Mythos AI and have not been independently validated.

Mythos AI has added partners for autonomous vessels and navigation hardware. In November 2025, [Ocean Power Technologies announced](https://oceanpowertechnologies.gcs-web.com/news-releases/news-release-details/ocean-power-technologies-and-mythos-ai-partner-advance?ref=runtimewire) plans to integrate Mythos AI's software with its WAM-V autonomous surface vehicles and PowerBuoy platforms, with initial demonstrations scheduled for the first quarter of 2026. No authoritative source cited here confirms that those demonstrations occurred. [BlackSea Technologies named Mythos AI](https://www.blacksea.tech/press/blacksea-technologies-names-mythos-ai-as-first-mainstay-autonomy-partnernbsp?ref=runtimewire) the first autonomy partner for Mainstay, a software architecture used across BlackSea's unmanned vessels, in July 2026.

Public disclosures do not establish Mythos AI's total funding, valuation, pricing or full institutional investor roster. PortXL has been identified as an accelerator or incubator backer, though no institutional round history is available in the cited materials. Mythos AI has disclosed more than 11,000 autonomous miles, while independent performance data for its APAS commercial deployment remains unavailable. That disclosure gap limits comparisons with better-capitalized road-autonomy companies and leaves the Southern Devall installation and partner integrations as its clearest public measures of commercial traction.

### Urmson pays for full autonomy

[Chris Urmson](https://ir.aurora.tech/company-information/leadership-team?ref=runtimewire) co-founded Pittsburgh-based Aurora Innovation in 2017 after leading Carnegie Mellon teams in the DARPA Grand Challenge and Urban Challenge and serving as chief technology officer of Google's self-driving program. His co-founders included former Tesla Autopilot director Sterling Anderson and Carnegie Mellon robotics professor and former Uber Advanced Technologies Group executive Drew Bagnell.

Aurora entered the public market in 2021 through [a business combination that produced more than $1.8 billion in gross proceeds and cash on hand](https://ir.aurora.tech/sec-filings/all-sec-filings/content/0001193125-21-319524/d249044dex992.htm?ref=runtimewire).

Aurora [reported launching a second-generation driverless fleet based on the International LT platform in late July 2026](https://ir.aurora.tech/sec-filings/all-sec-filings/content/0001828108-26-000075/aurora26q2shareholderlet.htm?ref=runtimewire). In its second-quarter shareholder letter, the company reported nearly 440,000 driverless miles through June 30, 100% on-time performance and no collisions attributed to the Aurora Driver. Aurora also said it expects to finish 2026 with more than 200 driverless trucks in operation. The mileage, performance and collision figures cover results reported through June 30, while the fleet count is a forward target for year-end.

Aurora's financial statements show the cost of moving from a technical demonstration to a freight service. In its [second-quarter 2026 shareholder letter](https://ir.aurora.tech/sec-filings/all-sec-filings/content/0001828108-26-000075/aurora26q2shareholderlet.htm?ref=runtimewire), Aurora reported $2 million in second-quarter revenue, a $266 million operating loss and approximately $225 million in cash used for operating activities. Aurora ended June with nearly $1.2 billion in cash and short-term investments. It expects $14 million to $16 million in revenue for all of 2026, with more than half projected for the fourth quarter. The revenue outlook is forward-looking.

Aurora is seeking to place trucks with customers through its Driver as a Service model. On April 30, 2026, the company [announced a nonbinding memorandum of understanding with Hirschbach Motor Lines](https://ir.aurora.tech/news-events/press-releases/detail/136/leading-carrier-selects-aurora-to-scale-autonomous-fleet-to-500-trucks?ref=runtimewire) under which Hirschbach would own and operate 500 trucks powered by the Aurora Driver, with deliveries beginning in 2027. Aurora said the contemplated arrangement included a Driver as a Service subscription, but final terms and binding truck orders remained subject to later agreements.

Kodiak AI, an autonomous-trucking developer, is already reporting traction with a customer-owned model. Kodiak AI said it had [28 customer-owned driverless trucks and more than 23,500 paid driverless operating hours](https://investors.kodiak.ai/news-releases/news-release-details/kodiak-ai-reports-first-quarter-2026-results-announces-100?ref=runtimewire) through March 31, 2026. Other competitors include Waabi, Plus, Gatik and Torc Robotics, a Daimler Truck subsidiary that is integrating its system with the Freightliner Cascadia and [targets commercial availability in 2027](https://torc.ai/frequently-asked-questions/?ref=runtimewire).

Urmson's utilization thesis requires roadside operations to keep pace with the driving software. Aurora says it has begun [supervised testing of weigh-station navigation and on-route fueling at truck stops](https://ir.aurora.tech/sec-filings/all-sec-filings/content/0001828108-26-000050/aurora26q1shareholderlet.htm?ref=runtimewire). Fueling, maintenance, loading and remote support will determine how much additional daily use Aurora can extract from each truck.

### Two models, different proof points

Established suppliers provide some context for Douglass's approach. Rail-equipment maker Wabtec expanded its inspection business in July 2025 by [paying $1.78 billion for Evident's Inspection Technologies division](https://ir.wabteccorp.com/news-releases/news-release-details/wabtec-finalizes-acquisition-evidents-inspection-technologies/?ref=runtimewire). Wabtec separately documents machine vision, sensors and AI-assisted analytics for rail inspection. Mythos AI is pursuing a smaller supplier model by fitting software to vessels its customers already operate. Aurora is building commercial operations around its driving system while pursuing conditional agreements for carrier-owned trucks.

Road autonomy also carries a regulatory burden because driverless trucks must respond to unpredictable public-road conditions. On July 8, 2026, the [National Highway Traffic Safety Administration warned automated-vehicle developers](https://www.nhtsa.gov/press-releases/av-developers-automated-vehicle-that-cannot-safely-interact-first-responders-danger?ref=runtimewire) after documenting vehicles entering emergency scenes, blocking responders or failing to recognize lights, flares, smoke, fire and traffic cones. The warning describes situations road-autonomy systems must handle before continuous operation can become routine.

Douglass can sell assistance before proving fully unmanned river operations. Aurora, under Urmson, reported second-quarter revenue that remained small relative to its operating loss as it expanded driverless service. Mythos AI now needs independently measured results from commercial vessels. Aurora must turn a growing fleet and hundreds of thousands of driverless miles into recurring revenue without compromising safety.
