The Defense Innovation Unit (DIU) has advanced Forterra and May Mobility to Phase III of its Robotic Operation for Autonomous Delivery and Sustainment (ROADS) prize challenge, moving the team toward an on-installation robotaxi pilot for U.S. Army Soldiers. The two companies announced the selection on November 5, 2026, stating they were chosen alongside four other vendors. ROADS is an up to $30 million effort to test and adopt commercial self-driving vehicles, including base shuttles and logistics trucks, across the Department of War’s non-tactical vehicle fleet.
DIU and the Army launched ROADS on May 22, 2026, targeting a fleet of more than 150,000 non-tactical vehicles (NTVs) that DIU says lacks modern dispatch, coordination, and automation tools, driving low utilization, excess capacity, and avoidable accident exposure. The cost is also measured in Soldier hours, as personnel are pulled from training to drive troops and equipment across post. The challenge runs in three phases: proposal evaluation of up to 10 companies, live demonstrations with awards of up to $500,000, and a Phase III on-installation pilot in which selected companies are eligible for up to $10 million.
How Forterra and May Mobility Divide Roles in the DIU ROADS Challenge
The team pairs a commercial autonomy developer with a defense fleet integrator. May Mobility supplies its patented autonomous driving system (ADS), integrated on passenger vehicles built for autonomy, which Soldiers would request through a rideshare-style app if the service is adopted. Forterra acts as fleet operator during Phase III prototyping, validating and testing the service through OneLink, its fleet command-and-control system. According to Forterra, OneLink gives installation leaders a live map and dashboard of every vehicle and lets them set installation-specific operating parameters and requirements.
May Mobility’s stack rests on a predictive world model and a real-time reasoning engine. The system simulates up to 10 seconds into the future, evaluates multiple driving strategies, and executes the one best suited to the situation. Publicly available information indicates its commercial fleet relies largely on Toyota Sienna minivans adapted for autonomy. The company reports more than 550,000 commercial autonomous rides across 1.1 million miles in the U.S. and Japan, including driver-out service, with partners including Toyota Motor Corporation, Lyft, and Uber.
Forterra, formerly known as Robotic Research Autonomous Industries, brings the military side. The company says it has fielded more than 250 autonomous systems for Department of War programs and has already run robotaxi service for Soldiers at Womack Army Medical Center at Fort Bragg, North Carolina. Phil Cotter, Forterra’s Head of Growth, Sustainment and Shaping Operations, linked the bid to the company’s deployment in Ukraine, describing the aim as bringing autonomy “that’s already working in the real world” to Service Members in garrison.
Fort Benning Pilot Sets the Phase III Evaluation Criteria
Vendor mapping begins in November 2026, followed by phased demonstrations and development through the end of the year, including a month-long pilot at Fort Benning, Georgia. DIU and the installation will assess technical performance, operational suitability, user needs, safety, and data management to inform the program’s next phase and any broader adoption across the department.
Safety is layered through onboard safety operators, remote assistance, and conservative safe-stop behavior, tracking the solicitation’s technical bar. DIU requires vendors to demonstrate, or show a credible path within 12 to 24 months to, SAE J3016 Level 4 operation with teleoperation fallback for edge cases, on widely available commercial base vehicles such as sedans, SUVs, pickups, or minivans. Vehicles must handle paved roads and well-maintained unpaved surfaces such as engineered gravel; no dedicated off-road capability is required.
Key details remain undisclosed. Neither the companies nor DIU have identified the four other Phase III vendors, the fleet size or vehicle platform planned for Fort Benning, or the value of any Phase III award to this team. It is also unclear whether safety operators will remain aboard for the full pilot.
Analysis: ROADS tests acquisition mechanics more than technology. Commercial robotaxi stacks already operate on public roads; the open question is whether the department can absorb them through a prize authority and move to follow-on production rather than stall in prototyping. The Forterra and May Mobility pairing reflects a model likely to recur, with commercial firms supplying the driving stack while defense-native integrators own fleet command and control and the safety case on government property. Garrison logistics also give the Army a low-risk setting to build familiarity with autonomy before relying on it in contested logistics.
The next markers are the completion of vendor mapping, demonstrations through December, and the Fort Benning pilot results. Based on Phase III performance, the government may scale selected solutions across multiple installations or keep maturing them with industry. Whether DIU names the full Phase III field, and whether driver-out operations are authorized on post, will signal how quickly the department intends to move.
