Large stealth drone without a tailplane, similar to a J-10, seen in Beijing: DSI intakes, EOTS, internal weapons bay, supersonic speed, UCAV mission, and collaboration.
The September 3 parade in Beijing featured a large stealth drone without a tail, comparable in size to a J-10. Its modified delta wing, DSI air intakes, serrated nozzle, and EOTS sensor under the nose indicate supersonic performance, an internal weapons bay, and an unmanned air superiority mission. A more compact Type A was also on display. The whole thing points to a Chinese strategy combining high-performance UCAVs and human-machine collaboration with fighters such as the J-20/J-20S. The exact status (advanced model or flight-capable airframe) remains to be confirmed, but the consistency of the structural details, the preliminary imagery and the markings suggest a serious program with a clear message: the rise of combat drones capable of carrying missiles internally and operating in a network.
The context of the parade and the rise of Chinese UCAVs
The September 3 parade in Beijing, marking the 80th anniversary of the victory over Japan, served as a showcase for unmanned systems and China’s doctrine focused on combat mass and survivability. In addition to ballistic vectors, several combat drones were displayed on truck carriers, without tarpaulins, revealing their detailed geometries. Among them, a large stealth drone, designated by observers as Type B or UASF (Unmanned Air Superiority Fighter), stands out for its “fighter” size, similar to a J-10 (length ~16.9 m, wingspan ~9.8 m). The slimmer Type A also adopts a tailless configuration, but with a lambda wing and caret-style air intakes, without a prominent EOTS bubble.
This Type A/Type B duo fits in with an already visible trajectory: China has multiplied iterations of UCAVs (e.g., GJ-11 Sharp Sword, flying wing) and is working on naval compatibility, internal munitions carriage, and electronic attack. The strategic signal from the parade is twofold. First, the PLAAF is promoting a complete range from collaborative wingman to autonomous UCAV, with air-to-air and air-to-ground roles. Second, the aerospace industry is demonstrating its ability to rapidly integrate everything from sensors to propulsion systems, with aggressive communication about technological sovereignty.
The operational consequence: in the short term, even pre-production airframes are used to train development and test chains (sensors, data links, thermal management, RCS). In the medium term, the objective is to create swarms and composite packages combining J-20/J-35, CCA, and UCAV to saturate the enemy’s defenses, stretch its radars, and impose a higher air tempo than that of Western forces.
Type B design: a stealthy, supersonic air-to-air profile
The Type B adopts a tailless airframe, with a modified delta wing in a diamond shape with truncated tips and large control surfaces on the trailing edge. The fuselage has a pronounced chine connecting the lateral DSI air intakes. This choice limits the complexity of the boundary layer separators, reduces the number of critical parts and, above all, indicates a supersonic objective. The DSI generates a stationary shock wave that is useful for controlling flow at high Mach numbers and eliminates the need for movable ramps, resulting in greater mass and reliability.
The presence of a variable geometry nozzle (probably afterburner) with serrations on the vein and fairing suggests partial treatment of the rear RCS. This compromise—performance vs. rear discretion—is common: the F-35 retains a thermally and mechanically optimized circular section, at the cost of a slightly more observable rear. Future iterations of the Type B could incorporate a 2D nozzle with thrust vectoring, improving signature and maneuverability at high angles of attack.
Under the nose, a large EOTS suggests IRST, laser telemetry, and multispectral imaging capabilities. Coupled with secure data links, the EOTS is used to detect and sort long-range targets, particularly in jamming environments where the AESA radar of a partner fighter may be turned off. The hatches visible under the airframe and the available volume suggest an internal weapons bay sized for medium/long-range air-to-air missiles and, depending on the configuration, air-to-ground munitions.
In terms of energy, a modern single-engine can provide thrust of around 80 to 120 kN in afterburner mode (indicative values depending on engine family). At Mach 1 at sea level, the speed is approximately 1,225 km/h (lower than at altitude). In an interception mission, an UCAV without a pilot or ejection seat reduces the weight by several hundred kilograms, freeing up fuel or payload in a volume comparable to a J-10.
The target mission: unmanned air superiority, distinct from CCAs
Unmanned air superiority differs from a cost-optimized CCA. A UCAV such as the Type B focuses on speed, penetration, and resistance to enemy sensors to intercept, lock onto, or divert enemy fighters, or even impose favorable firing windows for friendly platforms. The unit cost will be higher than a basic loyal wingman drone, as the airframe, RAM coatings, engines, and avionics must meet the same specifications as a light fighter. However, the absence of a pilot reduces training costs and saves scarce human capital.
A useful comparison: the Turkish Bayraktar Kizilelma, a delta-wing canard with a single engine, is designed to be supersonic (later versions), with an emphasis on thrust-to-weight ratio and payload rather than advanced stealth. Conversely, Type B favors frontal stealth and internal fuel tanks, at the probable cost of greater complexity and cost. In an Indo-Pacific theater, a stealthy supersonic UCAV can open corridors for long-range air-to-air missiles, cut air logistics lines, and force the adversary to turn on its radars, increasing its exposure to anti-radiation weapons.
Consequences: if the PLAAF manages to industrialize the Type B even in small batches, it will gain a sophisticated wear-down capability. An initial batch of a few dozen units capable of operating in pairs or trios with J-20/J-35 would be enough to expand the combat volume, make the enemy’s air defense more costly, and increase psychological attrition on the enemy pilot’s side, who knows they are being targeted by unmanned vehicles ready for high-risk kinetic exchanges.
Instructions for use: collaboration with J-20/J-20S, sensors, and autonomy
The two-seater J-20S has often been cited as a potential mission commander for combat drones, with the rear crew member managing communications, data fusion, designation, and deconfliction. In this scenario, a discreet Type B in front detects using IRST/EOTS, transmits tracks, and receives fire authorizations or no-fly zones. In the presence of jamming, sensor diversity (EO/IR + radar + ESM) improves the resilience of the whole system.
The next step is autonomy, with a tactical autopilot constrained by rules of engagement: compliance with geofencing, corridors, identification thresholds, and firing parameters. China clearly accepts a higher degree of lethal autonomy than the United States in the public domain, which speeds up testing. But this adds risks: imperfect understanding of a multi-sensor situation, friendly fire if the IFF is deceived, and unintended escalation. Technically, multi-hypothesis fusion algorithms, broadly trained classifiers, and massive air-to-air combat simulators are essential to achieve reliable decision rates.
In terms of armament, a bay capable of carrying air-to-air missiles measuring 3.2 to 4.0 m would impose constraints on the size of the payloads and the ejection kinematics. An internal rail or a rotary launcher imposes choices on the compatibility of homing devices and thermal management. Autonomy also implies careful fuel and thermal management: at high Mach, the flow over the skin heats up, which complicates the infrared signature. A RAM coating and carefully designed leading edges limit X/S band detection, but VHF/UHF radars remain a challenge, hence the interest in multi-axis approaches and low routes when the mission requires it.
Industrial and strategic implications: real maturity or theater?
The visible numbers, assembly quality, access panels, and details near the landing gear suggest realistic airframes. But a parade also serves as a political and industrial tool. It is possible that Type A and Type B are competing demonstrators from different teams, with selection to follow after flight testing. This is a classic process for reducing technical risks through internal competition.
The contrast with the United States is clear in the public arena: Washington is investing mainly in less expensive CCAs and agrees to little communication on a UCAV of superiority. In plain language: China is advancing rapidly in the stealth UCAV segment, while visible American programs have abandoned this niche since the X-45/X-47 demonstrators. As a result, even if the PLAAF does not immediately industrialize, it is accumulating feedback at a steady pace, while the USAF is focusing primarily on supporting manned fighters with simpler drones.
On the industrial front, the shift to domestic engines, sensors, and encrypted internal links reduces vulnerability to external supply chains. In budgetary terms, a high-performance stealth UCAV will remain expensive: coatings, quality of composites, tight tolerances, test benches for signature and vibrations. But the overall cost of ownership can remain competitive if pilot training is avoided, fighter airframe hours are saved, and higher equipment attrition in the field is accepted. This is the frank thinking that comes through: China is not trying to be “reasonable”; it is seeking useful military effect, even if it means cutting back on short-term financial elasticity.
War Wings Daily is an independant magazine.