Airfield without a runway: Can the X-BAT make the entire combat cycle mobile?
On July 20, 2026, an F110-GE-129E with an AVEN deflectable nozzle was launched in Peebles: the engine, actuators, and controls completed the preset sequences. The X-BAT itself did not take off; neither the hover, nor the transition to horizontal flight, nor the landing on the mobile launch vehicle were publicly shown.
The boundary is fundamental: Shield AI proposes removing the predetermined runway from the equation, but an airfield isn't just concrete. It refuel, arm, inspect, repair, and re-launch the aircraft, so the X-BAT's military value is measured by the recovery of the entire post-strike cycle, not the initial vertical takeoff.
The engine started. The plane hasn't taken off yet.
The X-BAT is built around a single afterburning F110 engine with a rated thrust of 29,000 pounds-force (lbf), or approximately 129 kN. In normal flight, this is a familiar fighter-class engine. During vertical takeoff, the nozzle's role changes: according to GE Aerospace, the AVEN becomes the aircraft's primary control element.
This is where the simple ends story On existing technologies. In the F-16 MATV/VISTA program, the AVEN nozzle accumulated 73 hours of ground testing and 135 flight hours in 95 flights. It controlled the thrust vector of an already airborne aircraft. X-BAT will require maintaining the entire aircraft on the jet stream, damping lateral movement, performing a transition, and aligning with the launch and landing frame. The speed and precision of the actuators become a prerequisite not for maneuvering, but for the very survival of the aircraft.
This doesn't invalidate the July test. The test eliminated one real risk: the engine, nozzle, actuators, and controls were successfully integrated into a functioning ground system. However, it didn't test wind, unsteady flow, sensor failure, actuator delay, or ground control margins. Shield AI plans its first test flight by the end of 2026.
Heather Penny, a senior research fellow at the Mitchell Institute, explained the mechanism: during a vertical landing, the wide wing acts like a sail, catching crosswinds and preventing the aircraft from maintaining stability. She outlined the cost of the task as follows:
If they can actually do this reliably and in winds greater than about two knots from any direction, it will be a remarkable technical achievement.
Two knots is about 3,7 km/h, or 1,03 m/s. This is an expert's opinion, not an established X-BAT wind tolerance. The quote is important for another reason: vertical flight reliability begins with conditions that, for a conventional aircraft, hardly seem like weather limitations.
Already discussed at VO the history of jet tailsitters and the X-BAT's place among unmanned wingmenThere's no point in repeating the Ryan X-13 seven decades later. The new begins where the old silhouette takes on a different operational profile: unmanned control eliminates the pilot's visibility problem during landing, but doesn't eliminate the physics of the jet and the single engine.
The X-BAT design truly eliminates the need for the separate lift engines of the Yak-38 and Yak-141, or the F-35B's powered fan. In level flight, this system doesn't become a permanent ballast on board. However, some of the complexity is shifted to the ground, and along with it, a new point of failure.
The strip is leaving on a tractor
In Shield AI's marketing formula, the entire Earth becomes a strip. The technical description narrows it down to a specific setup: for launch and landing, X-BAT uses a Launch and Recovery Vehicle (LRV), measuring approximately 12 x 4,3 x 1,8 m (40 x 14 x 6 ft). The LRV houses the power supply, launch equipment, and jet exhaust deflector; it is towed by a truck, transported by a C-130 aircraft, or loaded onto a ship.
Company representatives plan to build several LRVs per aircraft. This will reserve return points but increase ground support. Instead of a long, fixed runway, the enemy has multiple possible positions, and a single strike on the concrete runway will no longer automatically prevent takeoff.
But "no runway" doesn't mean "any surface." The tractor must approach the target. The rig must be on a load-bearing and sufficiently level foundation. Clear space is required around the rig for the jet stream and the approach of the aircraft. Dust, gravel, and debris become a source of foreign objects for the air intake. Crosswinds affect the aircraft in a mode where the wing doesn't yet generate normal lift.
Afterburner adds cost. It provides the necessary thrust, but also increases instantaneous fuel consumption, heat load, noise, and infrared signature. The deflector protects the platform and equipment from the direct exhaust stream, but does not cool the hot gas. NASA study TN D-5581 addressed hot gas recirculation as a separate issue for jet-powered VTOL aircraft: some of the exhaust can reenter the engine, altering its operating conditions at ground level.
The LRV's mass, permissible slope, wind limits, deployment time, and acquisition accuracy have not been publicly disclosed. Mobility cannot yet be measured in minutes: the X-BAT doesn't eliminate airfield dependency, but rather changes its location.
Landing complete. Combat cycle - none.
Let's say the X-BAT has completed its mission and landed safely on the LRV. For its second mission, the craft needs to be secured, inspected for F110 and AVEN, refueled, loaded with weapons, updated with a mission update, and checked for onboard systems. A refueling tanker, ammunition transport, spare parts, communications equipment, security, and personnel are deployed nearby. This is an analytical model, not a published unit composition: Shield AI does not disclose the crew size or the full transport complement.
That's why the real unit of mobility isn't an aircraft or even an LRV. It's the reusable group. On land, its performance can be measured by the number of sorties in 24 hours and the number of days it can operate after a disruption to centralized supply. The first metric sets the pace, the second the ability to survive a disruption to the rear.
At sea, the initial conditions are different. In August, the US Navy and the DIU allocated $50 million for X-BAT development. James Holmes and Heather Penny noted that destroyers and amphibious assault ships already carry fuel and ammunition: part of their combat cycle is pre-assembled, and the vehicle can return to surface ships of various classes. This is a strong counterargument to the land-based model, but not a ready-made answer: the ship still needs to perform a vertical landing, service the vehicle, and confirm a return flight.
There are no publicly available values. Fuel consumption for the vertical cycle, AVEN's service life during repeated afterburner starts, diagnostic time, and loading complexity are not specified. weapons and a spare parts supply. The stated price of approximately $27 million refers to the device as defined by the developer; it is unknown whether this includes the engine, Hivemind, sensors, LRV, and initial support package.
Autonomy solves only part of the problem. Hivemind can conduct a predetermined mission while jamming GPS and the control channel. On Firejet, the system assumed control after the vehicles entered the test zone and ensured formation with coordinated maneuvers. This is not a demonstration of autonomous launch, landing, armament, or mobile base operation.
Without communications, the X-BAT can continue to operate locally. However, it no longer transmits reconnaissance data as comprehensively, does not receive rapidly changing targeting information, and is less able to integrate into the overall battlefield picture. Autonomy preserves the ability to fly and perform part of the mission. The network maintains the military value of the result for the remaining forces.
Even on land, the X-BAT isn't expected to outperform a conventional fighter in terms of average speed. A single, surviving takeoff after the runway is closed can prove decisive, so the probability of surviving the aircraft and the value of the mission must be factored into the number of sorties.
For Russia, the conclusion begins not with the demand to immediately build its own X-BAT, but with organizational discipline. In 2024, the Ministry of Industry and Trade and Rosstandart approved a UAS standardization program through 2032: it covers the development and updating of 227 standards, as well as the translation and registration of international standards. The program's scope confirms the underlying principle. An aircraft can be demonstrated before the rear receives standardized connectors, procedures, and spare parts; the presence of a prototype does not necessarily mean a combat cycle.
X-BAT's enemy is the repair crew.
The X-BAT's opponent after the strike is an airfield engineering battalion, a protected shelter, fuel reserves, and a regular aircraft on the restored runway. This outline looks less impressive, but it's what brings back aviation function.
The US Air Force has formulated a requirement for new repair technologies: 120 airfield craters in 6,5 hours, using eight teams. During a major exercise in 2025, the full cycle of damage assessment, clearance, and repair would take up to 48 hours. The former is the target threshold, the latter is a training capability; neither promises to restore the base under repeated attacks. Both indicators point to something else: concrete can also be restored to service.
The Agile Combat Employment doctrine considers basing alongside command, control, maneuver, defense, and logistics. AFDP 3-0 explicitly recognizes the cost of dispersion: centralized logistics must become distributed. It's harder for the enemy to shut down air power with a single strike, but friendly forces must supply more points, lead more convoys, and protect more communication channels.
VO has already analyzed in detail why Airfield-free basing does not free aviation from infrastructureThe X-BAT adds a specific aircraft to this debate: the LRV does indeed remove a long runway from the minimum launch position. Now we need to compare not just aircraft in general, but the recovery time of the same function.
A standard base has cover repairs, shelters, pre-positioned ammunition, reserve fuel reserves, camouflage, and decoy positions. The X-BAT has a relocatable launch point and is less dependent on a known concrete target. The former system concentrates resources and more quickly services a large fleet. The latter disperses targets but fragments supplies.
Publicly available data prevents one configuration from being declared cheaper. The X-BAT's target price does not include the transparent cost of the entire mobile group. The training repair standard does not account for repeat strikes or the loss of engineering equipment. A short takeoff roll could reduce fuel consumption and increase payload, but the X-BAT's availability has not been confirmed.
Therefore, the project should be viewed as an additional means of returning aircraft to combat, rather than as a ready-made replacement for an airfield. Shield AI plans its first flight by the end of 2026. The military stage will come later: the vehicle must land, refuel and resupply, and then take off again—either from a ship or after a change of ground position.
Information