Before astronauts set foot on the Moon again, fleets of autonomous robots will need to build the port. Landing pads, roads, berms, blast shields, and the foundations for a nuclear reactor — the groundwork for a permanent lunar base will be laid not by human hands, but by robotic machines running on regolith and software, years before any crew arrives. That's the core premise behind Astroport Space Technologies , a San Antonio-based firm that just proved its approach works. In a test recorded and released this month, an Astrolab FLEX rover approached, attached, and operated Astroport's cylindrical drum excavator to scrape up layers of simulated lunar regolith. The demo moved the technology out of the theoretical and into the operational, and it comes at an accelerating moment for lunar infrastructure planning. AI-generated image Astroport's drum excavator attachment mounted to a FLEX rover, tested on simulated lunar regolith in March 2026. Credit: AI-generated illustration The Infrastructure Problem Nobody Talks About When a rocket lands on the Moon, it doesn't touch down on a groomed runway. It descends into loose, jagged regolith — gray dust with razor-sharp particles that have never been smoothed by wind or water. The rocket's exhaust plume, at velocities that can exceed 3,000 meters per second, blasts that regolith outward in every direction. Rocks and particles become projectiles, threatening any hardware within several hundred meters, including landers, rovers, and eventually habitats. This is called the lunar ejecta problem, and it's one of the reasons NASA's own safety panels have flagged the surface infrastructure gap as one of the more underappreciated risks in the Artemis program. There are no paved landing pads on the Moon. No roads. No blast shields. And every mission that touches down in bare regolith creates an ejecta hazard for the next one. The Ejecta Risk A single Starship HLS descent produces enough high-velocity regolith spray to damage hardware sitting 2 kilometers away. Without engineered landing pads, the lunar south pole's concentration of landers, rovers, and instruments becomes progressively more dangerous with each successive mission. NASA's Fission Surface Power program has set a 2030 target, backed by a December 2025 executive order, for delivering a nuclear reactor to the lunar surface. A reactor site requires a prepared foundation, cleared of loose material, leveled, and surrounded by protective berms. The window between now and 2030 is not wide. Robotic site prep needs to start well before any crew or reactor hardware arrives. Building the Port Before the Ship Arrives Astroport CEO Sam Ximenes describes his company's philosophy in a single phrase: "building the port before the ship arrives." The company, a subsidiary of Exploration Architecture (XArc) and based at Port San Antonio, designs interchangeable robotic construction tools meant to ride on commercial rover platforms. The tools include excavators, drills, trenchers, leveling blades, and shovels, each designed to be swapped out autonomously as the construction sequence demands. The partner for the mobility backbone is Venturi Astrolab , a Hawthorne, California company that built the FLEX rover, the largest and most capable commercial rover currently under development for the Moon. FLEX can carry up to 1,600 kg of cargo, transport two suited astronauts, and swap payloads through a standardized interface. The collaboration turns FLEX into what Astrolab CEO Jaret Matthews calls "the Swiss Army Knife of lunar construction." 1,600 kg FLEX rover cargo capacity 5+ Interchangeable construction tools $1.4M NASA STTR grant portfolio 2030 Target year for lunar nuclear reactor 15 Astroport employees across 3 continents 20 km/h FLEX rover max speed on lunar surface The March 2026 excavator test is the first step in a sequence. Astroport's plan is to send fleets of these construction rovers to the Moon ahead of crewed missions to prepare landing pads and launch pads, compact and grade roads between landing zones and habitats, construct regolith berms around sensitive equipment, and clear and prepare the site for the first lunar nuclear reactor. FLEX: The Platform That Makes It Possible AI-generated image Concept render of the FLEX rover operating at the lunar south pole. The rover's modular cargo interface accepts interchangeable tools for different mission phases. Credit: AI-generated illustration Venturi Astrolab launched FLEX as a commercial product to solve what its founders identified as the lunar last-mile problem. Landers can deliver payloads to the surface, but moving them, deploying them, and maintaining them over time requires a capable surface vehicle. FLEX is designed to be that vehicle, and its commercial payload interface, with dust-tolerant quick-disconnects for power, data, and thermal connections, is what makes the Astroport partnership viable. The rover uses a wheel-on-limb suspension system adapted for the Moon's soft regolith, with four wide airless wheels that can handle crater rims, slopes, and loose soil. It operates semi-autonomously from Earth and can also be controlled directly by astronauts. Power comes from a 3-square-meter deployable solar array, with Venturi's batteries rated for the extreme temperature swings of the lunar south pole, from near absolute zero in shadowed craters to over 100°C in direct sunlight. Astrolab is also developing a smaller technology demonstrator called FLIP, which launches to the lunar south pole aboard Astrobotic's Griffin-1 lander (on a Falcon Heavy) as early as summer 2026. FLIP will test FLEX's core subsystems, wheels, batteries, avionics, and dust mitigation, in actual lunar conditions before the full FLEX platform makes its first commercial mission. FLEX Rover Specifications • Cargo capacity: Up to 1,600 kg with 3 cubic meters of cargo volume • Robotic arm: 6 degrees of freedom, 25 kg payload, 2-meter reach • Crew capability: Transports 2 suited astronauts in standing configuration • Autonomy: Hazard avoidance, semi-autonomous navigation, teleoperable from Earth • Power: 3 m² deployable solar array, high-cycle batteries for south pole operations • Interface: FLEX Universal Payload standard, dust-tolerant quick-disconnects The Material Is Already There One reason the Astroport approach has attracted sustained NASA funding, including multiple SBIR and STTR contracts totaling $1.4 million across 2024 and 2025, is the fundamental logic of using the Moon's own material as construction feedstock. Shipping construction supplies from Earth costs roughly $1 million per kilogram to the lunar surface under current pricing. Lunar regolith, by contrast, is everywhere and costs nothing to acquire. AI-generated image Artist concept of the BrickBot system sintering lunar regolith into construction bricks. Astroport has demonstrated key milestones in this process through 2025-2026. Credit: AI-generated illustration Regolith can be used in multiple ways. Loose and compacted, it works as a substrate for landing pads and roads, essentially using the Moon's own material as gravel and fill. Heated to around 1,000°C, regolith particles fuse together in a process called sintering, forming solid tiles or bricks without any added binder. These sintered bricks have compressive strength comparable to concrete and can be laid by robotic systems to create durable paved surfaces. Astroport's BrickBot system handles this second process. A separate NASA STTR Phase II award ($849,994, active through January 2027) covers end-to-end 3D construction printing from raw lunar regolith, including excavation, filtering, melting, and forming and emplacing brick tiles for landing pads, roads, and planar structures. Milestones achieved through early 2026 include a working regolith conveyance system, a filtration and beneficiation stage, and integration testing with a brickmaking furnace prototype. Construction Method Regolith Use Application Status Compacti