Ukraine's Brave 1: The Autonomous Interceptor Built to Hunt Jet-Powered Gerans

Screen grab from Militarnyi video on YouTube embedded below

Ukraine is all set to field its first autonomous interceptor drone — the Brave 1. So far, Ukraine has developed and deployed FPV interceptor drones only.

The Brave 1 has successfully undergone combat tests in the Kharkiv region with a unit of the 12th Special Forces Group.

A video posted on the Militarnyi YouTube channel allegedly shows the drone intercepting an unidentified Geran variant and several other drones.

Brave 1 Features

The tube-launched Brave 1 features an aft-mounted pusher propeller and an optronic sensor in the nose.

The drone has a conventional airframe, making it aerodynamically more efficient than the Sting. The latter, which currently serves as the workhorse interceptor drone for Ukrainian forces, is an FPV drone with a bullet-like fuselage and a quadcopter architecture. The drone was operationally fielded around mid-2025.

Quadcopter drones have limited maximum speed because of the higher drag associated with the design. This limitation makes them ineffective against high-speed, jet-powered Geran one-way attack drones.

The conventional airframe of the Brave 1 would allow it to achieve much higher speeds. It is likely that, in a dive, it would be capable of intercepting a jet-powered Geran-4 or Geran-5.

Another limitation of the Sting is that it is an FPV kinetic interceptor that must be piloted into its target by a ground-based operator using an RF communication link. Signals from this link can be detected by adversary EW sensors, and the location of the pilot can be extrapolated through triangulation.

Russian forces reportedly exploit this vulnerability by pairing Geran-2 drones. When a Sting interceptor is launched against one Geran-2, a second Geran autonomously uses onboard sensors to locate and dive onto the Sting operator.

Interception Autonomy

The next generation of interceptor drones is being designed to operate autonomously using AI-powered machine vision, both during the day and at night.

These drones can independently detect, track, and engage targets without continuous human involvement. As a result, they do not require a vulnerable control channel that can be jammed or spoofed. They can home in on targets even when SATNAV and communication signals are jammed

In effect, they are launch-and-scoot weapons that remove the operator from the battlefield and significantly reduce operational risk.

Merops AS-3

Besides the Sting, Ukrainian forces use the US-supplied Merops AS-3 drone to intercept Geran drones.

The Merops AS-3 Surveyor is a mobile, truck-portable counter-drone system comprising:

1. Radar and electro-optical sensors for detection and tracking
2. A ground control and command station
3. Pneumatic or mobile launch platforms
4. A fleet of Surveyor interceptor drones

The AS-3 features a conventional fixed-wing airframe and AI-based autonomy. The drone is cued and initially guided using the sensors of the Merops system. For terminal guidance, it uses onboard IR and RF sensors, as well as AI-based machine vision. 

With a maximum speed of 280 km/h, the drone outpaces Russian Gerans more easily than the Sting.

Although highly effective, the Merops system relies heavily on ground-based infrastructure that can be located and attacked. Also, the system is expensive. It is currently priced at around $15,000.

Russian Yolka Interceptor Drone

At the beginning of the year, Russia fielded its Yolka interceptor drone, which featured the autonomy of next-generation drones and stole a march on the capabilities of the Ukrainian Sting.

The Yolka uses a hybrid quadcopter configuration augmented by X-planform lift-generating wings.

The Yolka is remarkably simple to use. It is launched from a handheld pistol-like device. The operator points the drone toward a target and launches it as soon as the drone begins to autonomously track the target. The UAV is fitted with electro-optical (daylight camera) and infrared sensors, along with an onboard AI processor known as the "Igolka" module.

The Yolka's portability, ease of use, and very low cost (approximately $500) enable widespread and distributed deployment.

Conclusion

Nothing definitive has been published about the guidance architecture of the Brave 1. However, since the drone is launched vertically from a tube, it is likely initially cued to its target by a ground-based radar sensor, as is the case with the Merops system. Mid-course guidance likely relies on AI-powered fusion of ground-radar and onboard optical-sensor inputs.

It is likely that Ukraine has developed the Brave 1 interceptor primarily to counter the threat posed by Russian high-speed jet-powered drones such as the Geran-3, Geran-4, and Geran-5.

The Geran-3 features a Geran-2 fuselage fitted with a jet engine.

The Geran-4 is an adaptation of the Geran-2 airframe optimized for jet propulsion. It features a compact jet engine mounted beneath the aft fuselage. The design is aerodynamically more efficient, resulting in greater range and speed.

The Geran-5, however, is a clean-sheet design. Russian forces first used the drone in January. The catapult-launched drone resembles a subsonic cruise missile, with a tubular fuselage and fold-out wings.

Ukraine’s Main Intelligence Directorate believes the Russian Armed Forces are actively ramping up their inventory of jet-powered drones to an extent where 50 percent of the long range one way attack drones launched at Ukrainian targets will be jet powered. 

In an earlier post, I discussed the Molniya, a second-generation Russian autonomous interceptor drone that retains the operational simplicity of the Yolka interceptor while offering greater capability through longer range, higher speed, and a warhead. It can reach speeds of up to 330 km/h.

One possible shortcoming of the Molniya is its quadcopter configuration. However, Russia may have deliberately chosen this design because it is also simultaneously developing the Molniya-P interceptor drone with a conventional airframe.

With the Brave 1, Ukrainian forces may well have narrowed the gap in interceptor-drone capabilities between the two nations thanks to its conventional airframe and the higher speeds that configuration enables.

In Ukraine, Cornered by US Supplied Hornet Drones, Russia Bounces Back with Molniya Purpose-Built Hornet Hunter

Molniya Interceptor Drone Undergoing Factory Testing. Screen grab from RuMoD released video.

Russian forces appear to be reeling under a sustained campaign of mid-range interdiction attacks on their logistics network by Ukrainian forces using US-made Hornet AI-powered autonomous attack drones. The Hornet can navigate using machine vision and autonomously hunt, recognize, prioritize, and attack targets moving along Russian supply routes.

As a result, over the past month, the flow of fuel, ammunition, and reinforcements to the front line is drying up.

Logistics Lockdown

Ukraine has succeeded in imposing a logistics lockdown on Russian forces using Hornet drones, a lockdown that appears to have significantly slowed the Russian offensive along the Donbas frontline..

It's conceivable that supplies reaching the frontline are barely allowing Russian forces to hold territory.

Breaking Out of the Lockdown

To retain the offensive capability of its forces, the Russian military leadership has taken immediate measures such as shortening convoys and avoiding highways by using alternative routes and dirt roads.

Short-term measures being considered to ease Ukraine's "logistics lockdown" include relocating depots deeper within Russian territory and redeploying short-range mobile air-defense systems, such as Tor and Pantsir, to protect highways from drone attacks. However, when resources are limited, strengthening the rear can weaken the front.

Relocating depots deeper within Russian territory will slow the flow of supplies. Redeploying air-defense assets will weaken the integrated air-defense network and make forward-deployed artillery and command posts more vulnerable.

Long Term Measures

It is the long-term measures that the Russian leadership is likely betting on, including 

1. Improving drone detection and tracking through the use of more capable radar and optical sensors. 

2. Deploying large numbers of interceptor drones to engage attacking Hornet drones.

Yolka Interceptor Drone

Yolka is a handheld, man-portable kinetic interceptor drone launched from a pistol-like device. It's extremely simple to use. The operator points it roughly toward a target, launches, and the drone autonomously tracks and rams the enemy UAV using electro-optical (daylight camera) + infrared sensors plus an onboard AI/processor ("Igolka" module). 

Weighing approximately 1.3 kg, the Yolka can reach speeds of 200–250 km/h and operate at altitudes of up to 2 km.

Russian forces began operational deployment of the autonomous Yolka interceptor drone in early 2026.

The Yolka's portability, ease of use, and very low cost (approximately $500) enable widespread and distributed deployment. 

The Yolka provides a credible counter-drone capability for mobile counter-drone ("drone hunting") teams, small infantry and special-forces units, and for protecting personnel, vehicles, and equipment in the field. 

Though the Yolka has been a success story, it was developed and operationally deployed before former Google CEO Eric Schmidt dug into his deep pockets to fund the development of the Hornet strike drone that is now causing anguish along the front line and at command centers. 

The speed and routing flexibility of the Hornet are well beyond the interception capabilities of the Yolka. 

Screen grab from RuMoD released video.

Molniya Interceptor Drone

Developed by the Scientific and Production Center for Unmanned Aviation Systems in Russia's Tomsk Region, the Molniya interceptor is somewhat similar in shape (bullet-like) to the Ukrainian Sting interceptor. Both use a quadcopter flight and control scheme.

Currently, Molniya is undergoing factory tests. 

Significantly, the Molniya is heavier (2.5 kg versus 1.3 kg) than the Yolka and offers a greater maximum engagement range (5 km versus 3 km).

Unlike the Yolka, which lacks a warhead and relies entirely on kinetic interception, the Molniya carries a 300-gram warhead. 

Both the Yolka and Molniya can be hand-launched and feature AI-powered autonomy that makes them resilient to control channel jamming. 

The Yolka starts using its optical / thermal sensors and AI interception algorithms from before launch in order to engage its target. The Molniya has optical and thermal sensors and can likely use onboard AI to fuse radar and optical sensor data for autonomous positioning and interception. 

While the Yolka needs to be visually cued onto its target before launch and cannot be effectively used under poor visibility conditions, the Molniya is cued by ground-based radar  facilitating use under all visibility conditions. Radar cueing also leverages the interceptor's longer engagement range making it more versatile than the Yolka. Early launch allows the drone to position optimally for an interception. 

Reusable Interceptor?

Perhaps its most intriguing feature is that the Molniya appears to be a reusable interceptor drone. It is equipped with landing struts that protect its propellers during landing. Footage of Molniya trials recently posted on social media shows the drone returning for a controlled landing. Indeed, the footage focuses almost entirely on the drone's return capability. 

For an interceptor drone to be reusable, it needs the ability to take down its target without destroying itself. Could the 300-gram warhead on the Molniya be ejectable, allowing it to return from a successful engagement? A close look at the drone video does not rule out the capability. However, most likely, the drone is just capable of aborting an interception. 

Screen grab from RuMoD released video.

Conclusion

Russia has a proven track record of countering sophisticated, versatile, and expensive Western weapon systems with less sophisticated, more focused, significantly lower-cost but effective alternatives.

Russia countered the Ukrainian deployment of the Sting interceptor drone with the significantly cheaper yet highly capable Yolka interceptor. 

The Molniya's longer range, radar cueing, and autonomous sensor-fusion tracking capabilities are reminiscent of the US Merops AS-3 interceptor drone used by Ukrainian forces, the development of which was also funded by Eric Schmidt. Unlike the Merops, however, the Molniya appears to be designed for mass production and widespread deployment. If it proves capable of reliably countering the Hornet, it may only be a matter of time before Russian forces are able to regain the operational momentum necessary to resume offensive operations in Donbas.

The Next India-Pakistan Conflict Will Be Won by Drone Killers

 

The ability to defeat drones is increasingly proving to be more important than the ability to field them. That is perhaps the most important lesson emerging from the wars in Ukraine and Iran—a lesson Indian military planners need to factor into preparations for the inevitable next conflict across our western border.

According to General Oleksandr Syrskyi, Commander-in-Chief of the Armed Forces of Ukraine, most Russian Shahed drones and other aerial attack systems neutralized by Ukrainian forces are now being brought down by interceptor drones.

Not helicopter gunships, armed light trainers, directed-energy weapons (DEWs), low-cost air-defence missiles, or specialised anti-aircraft guns firing programmable airburst ammunition. No—just interceptor drones.

Operationally effective interceptor drones have been around for more than a year. It is time DRDO took note of them.

Interceptor drones are optimised for low cost and typically destroy their targets by ramming them. Some variants employ a small warhead to increase the probability of a successful interception.

Currently, the three most prominent interceptor drones operating in Ukraine are:

* Sting (Ukraine) * Merops AS-3 Surveyor (United States) * Yolka (Russia)

In the following sections, we examine their features, capabilities, and key differentiators.

FPV Interceptor Drones

Interceptor drones were initially FPV (First-Person View) drones that relied on real-time radio control and live video feeds.

Using RF sensors, it is possible to detect the communication link between a drone and its operator, allowing triangulation and pinpointing of the operator's location. The control link can also be jammed or spoofed using electronic warfare (EW) systems.

FPV drone operators face significant danger because they must remain relatively close to the front line to maintain a strong signal. This exposes them to artillery fire, snipers, counter-drone attacks, and detection through radio-frequency triangulation.

Russian forces exploit this vulnerability by pairing Geran-2 drones. If an FPV interceptor is launched against one Geran-2, the second drone can use onboard sensors to locate the operator and attack immediately.

Ukraine's Sting Interceptor Drone

The most successful Ukrainian-developed interceptor drone currently in service is the Sting. The low-cost drone (approximately $2,000–2,100) uses a quadcopter architecture and features a 3D-printed aerodynamic airframe shaped like a bullet.

The Sting is capable of speeds of approximately 280–343 km/h. It is designed primarily as a kinetic interceptor, with the operator steering it directly into the target. Guidance is provided through day and thermal cameras, with possible sensor fusion from radar systems.

Ukraine began employing the Sting in combat during the spring of 2025, with widespread deployment by June 2025. The first publicly documented success occurred in April 2025 when footage of a Sting downing a Shahed-type drone went viral.

Sting Interceptor Drone Photo: The Telegraph

Perhaps the most remarkable feature of the Sting is its ability to be operated remotely from hundreds or even thousands of kilometres away.

Like other FPV drones, the Sting maintains a line-of-sight link to a forward control station. However, this forward station functions primarily as a relay node, connecting the drone to its pilot via Starlink's low-latency internet network. As a result, the pilot can be located virtually anywhere in the world.

Ukrainian operators typically control the drone from hardened shelters. The architecture also allows Sting drones to be launched from unmanned surface vessels (USVs).

Autonomous Interceptor Drones

The next generation of interceptor drones is being designed to operate autonomously using AI-powered machine vision, both during the day and at night.

These drones can independently detect, track, and engage targets without continuous human involvement. As a result, they do not require a vulnerable control channel that can be jammed or spoofed.

In effect, they are launch-and-scoot weapons that remove the operator from the battlefield and significantly reduce operational risk.

Russian Yolka Drone

Russian forces began operational deployment of the autonomous Yolka interceptor drone in early 2026.

The Yolka can be hand-launched, enabling widespread and highly distributed deployment. Once launched in the direction of a target, it operates autonomously.

Weighing approximately 1.3 kg, the Yolka is also based on a quadcopter architecture similar to the Sting. It can reach speeds of 200–250 km/h and operate at altitudes of up to 2 km.

Yolka vs Sting

The Yolka is significantly lighter and cheaper than the Sting, with an estimated cost of roughly $500 compared to the Sting's $2,000-plus price tag.

However, these savings come with trade-offs. The Yolka's range is limited to approximately 2.5–4 km, compared with the Sting's estimated 25–37 km range. It is also slower than its Ukrainian counterpart.

Merops AS-3 Surveyor

In addition to the Sting, Ukrainian forces are employing the American-made Merops AS-3 Surveyor, a sophisticated but significantly more expensive interceptor system.

Unlike the Sting and Yolka, the AS-3 requires catapult launch, reducing deployment flexibility. However, its conventional fixed-wing airframe enables much higher aerodynamic efficiency and speed.

The truck-portable counter-drone system consists of:

* Radar and electro-optical sensors for target detection and tracking * A command-and-control station * Pneumatic or mobile launch platforms * A fleet of Surveyor interceptor drones

The AS-3 derives its effectiveness from a combination of AI-enabled autonomy, high speed, and resistance to jamming.

After launch, the drone is initially guided using the sensors of the Merops system. During the terminal phase, it relies on onboard infrared and RF sensors combined with AI-powered machine vision.

The drone can continue homing in on its target even when satellite navigation and communication signals are jammed.

AI-powered machine vision, combined with the fusion of infrared and RF sensor inputs, is central to the Surveyor's effectiveness.

With a maximum speed of approximately 280 km/h, the AS-3 is capable of overtaking Russian Geran drones.

The current unit cost is estimated at around $15,000, although this is expected to fall below $10,000 as production scales.

Quadcopter vs Fixed-Wing Interceptors

The Sting and Yolka are quadcopter drones. They are simpler and cheaper to manufacture but are aerodynamically less efficient because they lack wings to generate lift and glide efficiently through the air.

Consequently, quadcopters are not optimised for sustained high-speed flight.

To intercept faster fixed-wing drones, quadcopter interceptors often position themselves above the incoming target. At the appropriate moment, they dive, converting altitude into speed and enabling a successful interception.

Future Developments

Interceptor-drone development is currently focused on increasing speed through the adoption of fixed-wing designs such as the AS-3 Surveyor.

Reusability is another area receiving considerable attention.

Fixed-wing drones can achieve higher speeds in level flight and generally manoeuvre more efficiently during the terminal interception phase.

Russian forces have already begun mass deployment of fixed-wing interceptor drones, including a dedicated air-defence variant of the Molniya family known as the Molniya-PVO. The drone is reported to be capable of speeds between 220 and 330 km/h.

Like the Yolka, the Molniya can be hand-launched. Alternatively, it can be launched using a lightweight catapult.

Conclusion

Ukraine seized an early lead in interceptor-drone technology with the Sting. Russia has largely closed the gap with the rapidly evolving Yolka and is now introducing fixed-wing interceptors such as the Molniya-PVO.

One important point stands out. Starlink has given Ukrainian forces a low-latency communications advantage that Russia is unlikely to match for several years.

There is another lesson for Indian defence planners. Rapid advances in drone autonomy are being driven by access to high-performance AI semiconductors and resilient communications networks.

India remains a long way from sovereign access to either low-latency broadband networks comparable to Starlink or the cutting-edge AI chips needed to support the next generation of autonomous combat systems.

MQ-9B LoyalEye: The Rise of Unmanned Airborne Early Warning

 

MQ-9B LoyalEye AEW maiden flight

GA-ASI and SAAB are collaborating to develop a remotely piloted aircraft (RPA)-based airborne early warning (AEW) platform.

GA-ASI is equipping its MQ-9B Remotely Piloted Aircraft (RPA) with Saab-developed LoyalEye radar sensor to create a low-cost platform that will complement manned AEW platforms.

On May 19, GA-ASI flew its MQ-9B for the first time with two Airborne Early Warning (AEW) pods that will eventually carry the fixed antennas of the LoyalEye radar, providing persistent and cost-effective air surveillance.

In the past, SAAB has developed two AEW&C platforms — the Saab 340 Erieye and its successor, the Saab GlobalEye. The Erieye is mounted on the Saab 340 twin-turboprop regional airliner, while the GlobalEye is mounted on the Bombardier Global 6000/6500 ultra-long-range business jet. Both feature AESA radars with fixed antennas.

This MQ-9B maiden flight marks the start of a several-month test-and-evaluation phase, which will culminate in a full-capability demonstration later this year.

The joint AEW offering from Saab and GA-ASI will support a wide range of applications, including early detection and warning, long-range detection and tracking, and the simultaneous tracking of multiple targets. The system will operate both beyond line of sight and via satellite communication (SATCOM) connectivity.

The platform will facilitate defence against tactical air munitions, guided missiles, drones, fighter and bomber aircraft, and other threats.

Using an RPA instead of a manned platform for AEW has some unique advantages, including:

1. High loiter time
2. Aircrews are not put in harm's way
3. Lower acquisition cost
4. Lower operating cost

Loiter time could possibly extend to nearly 40 hours, compared with the approximately 8–11 hours achievable by manned aircraft.

Aircrew and radar operators would be positioned on the ground rather than on the airborne platform. The arrangement is not just safer; it is also cost-effective. Aircrew, radar consoles, and control systems do not remain platform-specific.

The estimated price of the MQ-9B LoyalEye will reportedly be in the $60–80 million range.

The operating cost of an RPA platform would be significantly lower than that of a manned platform.

China's WZ-9 Divine Eagle

China has been actively pursuing remotely piloted airborne early warning (AEW) capability.

Its Shenyang WZ-9 "Divine Eagle" is a high-altitude, long-endurance (HALE), jet-powered UAV designed specifically for AEW and counter-stealth roles.

The twin-boom platform, featuring high-aspect-ratio main wings and a forward canard-style horizontal stabilizer, is powered by a single WZ-9 turbofan engine.

However, the Divine Eagle differs conceptually from the MQ-9B AEW.

The 15-tonne Divine Eagle is believed to be the largest UAV ever built. It is estimated to be roughly 15 m long, with a 45 m wingspan, offers approximately 35 hours of endurance, and has a service ceiling of about 25 km (82,000 ft).

Chinese media suggest that the UAV will be used for a variety of missions, including early warning, targeting, electronic warfare (EW), and satellite communications.

Implications for India

The possibility of Pakistan acquiring an MQ-9B AEW analogue in the future should worry Indian defence planners. The platform is relatively inexpensive but likely capable.

The Pakistan Air Force (PAF) already operates the Saab 2000 Erieye AEW&C (Airborne Early Warning & Control) platform. An export variant of the Saab 340 Erieye, the Saab 2000 Erieye uses the Saab 2000 twin-turboprop regional airliner.

The PAF is one of the largest operators of the platform, with nine aircraft in service as of mid-2026.

It could be possible for Saab to supply the MQ-9B AEW itself or, if the US would not allow it, fit the LoyalEye radar to an alternative MALE platform such as the Bayraktar Akinci.

The lighter payload capability of the Akinci (1,500 kg), compared with that of the MQ-9B (2,500 kg), would make the fit tight once fuel, sensors, and weapons are factored in. However, the Akinci's twin turboprops (up to 2 × 850 hp) likely offer comparable or greater electrical capacity and redundancy.

Not Just for Export: Russia Advances the Su-75 for Its Own Air Force

 

Su-75 Prototype? at Zhukovsky

Vadim Badekha, General Director of the United Aircraft Corporation (UAC), told TASS on June 2, "The work on Checkmate is already at the stage of building a prototype."

According to Badekha, Russia is developing the medium-weight fifth-generation fighter both for use by Russian forces and for international customers.

Badekha noted that the single-engine Checkmate has a cost advantage over heavy twin-engine fighters such as the Su-57.

Earlier, in January, the Izvestia newspaper reported that the first flight of the Russian fifth-generation Su-75 single-engine fighter could take place in 2026.

On November 18, 2025, during an interview on Russia's state-run Channel One television station, Sukhoi chief test pilot Sergei Bogdan stated that the first flight of the Su-75 would take place in 2026.

"The aircraft is already on the shop floor, it is already being finalized, and there are already certain timelines. Therefore, with God's help, it should take place soon enough."

Su-75 Journey So Far

Russia's Sukhoi company first unveiled the company-funded, low-observable (LO), supermaneuverable, optionally manned, single-engine, lightweight (18-tonne), Mach 2 fighter at MAKS 2021.

The Checkmate was reportedly developed based on an analysis of the use of strike aircraft in Syria, which revealed that, for most tasks, the capabilities of heavy twin-engine aircraft are excessive.

The aircraft is reportedly being developed using digital-twin technology, reducing development time by five years.

The USPs of the fighter, designed using electronic modelling, include:

1. Low acquisition and operating costs
2. AI assistance for single-pilot operations
3. Ease of maintenance

Timeline Slippages Due to Redesign

On the eve of Dubai Airshow 2023, Rostec told RIA Novosti that preparations had begun in Russia for the production of the first examples of the light stealth fighter.

As per the timeline announced at Dubai Airshow 2021, the first prototype of the aircraft was expected to fly by the end of 2023.

A statement released by Rostec's press service explained that the timeline had slipped to accommodate design changes. The redesign, it said, "has significantly increased the competitiveness and commercial attractiveness of the domestic single-engine aircraft and reduced the technical risks of its development."

First Deputy Chairman of the Federation Council Defence and Security Committee, Viktor Bondarev, confirmed that, following the redesign, the documentation for the Checkmate fighter had been transferred to the manufacturing plant and preparations for the production of the first examples had begun.

"The pilot batch of Checkmate fighters is planned to be produced in 2026," he said.

Deputy Prime Minister and Minister of Industry and Trade Denis Manturov said: "I can say that over these two years we have collected requests, including making certain adjustments to the project so that it is maximally adapted to the requirements of customers interested in a single-engine aircraft. This applies to the layout, control systems, and aviation weapons. A great deal of work was done on the basis of the Checkmate originally presented here."

In February 2024, Rostec stated in a statement to TASS: "Under the Checkmate program, the United Aircraft Corporation of Rostec received feedback from potential customers. In addition to collecting additional requirements expressed by potential buyers, work was also carried out to optimize costs and analyze individual technical solutions. This made it possible to significantly increase competitiveness and commercial attractiveness, while also reducing technical risks in the creation of a domestic single-engine aircraft."

Su-75 at Zhuhai

Prototype Development

In February 2025, Russian media reported that KnAAZ (Komsomolsk-on-Amur Aviation Plant named after Y.A. Gagarin) planned to assemble two prototypes of the Su-75 Checkmate light tactical aircraft (LTA), one for static testing and one for flight testing.

Three Variants

At Dubai Airshow 2025, as if to reiterate its commitment to the Su-75 project despite the war in Ukraine and unconfirmed interest from potential foreign customers, Sukhoi unveiled a model of an unmanned variant of the fighter alongside the model of the manned variant.

The unmanned variant featured a reworked wing and a revised rear fuselage.

As of now, the Su-75 is proposed in three variants: single-seat, twin-seat, and unmanned.

In the past, Russian media, quoting the Federal Service for Military-Technical Cooperation (FSMTC), reported that Russia and Belarus were working together on the Checkmate project.

The approximate cost of the Checkmate has been estimated at $30 million. For comparison, the basic version of the F-35 costs $80 million or more.

Indian Interest?

In February 2026, responding to a query from TASS, a HAL official said the company was ready to work with Russia to produce new warplane models.

"We are very comfortable working with Russia. There are no problems on our part," said the HAL representative.

In the past, HAL collaborated with Russian aircraft manufacturers to assemble and locally produce fighter aircraft for the IAF, including the MiG-21, MiG-27, and Su-30MKI. However, joint development had never previously been considered.

Joint development offers the advantage of Indian ownership. Beyond securing IPR for the weapon system and fostering self-reliance, it ensures that projects are not subject to import restrictions arising from sanctions.

However, it may now be too late for HAL to participate in the design and initial production of the aircraft. The design is frozen and Russian officials have confirmed that the initial batch of Su-75 fighters will be manufactured in Russia to meet the requirement of Russian forces. 

The Su-75 Checkmate Option

The TASS report was non-specific in nature, but it may have alluded to a specific project such as the development of the Su-75 stealth fighter. Russia has previously indicated that it is open to joint-venture production of the Su-75 in a partner country. Such a venture would align perfectly with India's Make in India and Atmanirbhar Bharat initiatives.

Alongside helping to plug the emerging stealth gap, local manufacture of the Su-75 in India could generate billions of dollars in export revenue, much like the BrahMos missile program and potentially the Su-30MKI modernization program.

India's manufacture of the Su-75 would in no way compromise the AMCA program. The AMCA is a twin-engine medium-weight stealth fighter, whereas the Su-75 is a single-engine lightweight stealth fighter. The two serve distinct operational roles. A single-engine fighter is cheaper to acquire and operate, and the Su-75 would remain relevant to the IAF's requirements well beyond the induction of the AMCA.

India Cannot Win Tomorrow’s Wars With Yesterday’s Technologies

 

AI conceptualisation of a AI powered drone attacking a supply truck

The US currently leads the world in two critical military technologies — satellite-based low-latency internet and AI. The former gives its weapons global reach, while the latter provides unprecedented accuracy. Together, they could enable the US to maintain its military dominance across the world for decades.

In a low-key manner, the US is already flexing its Starlink-based global reach and AI-powered accuracy through the drones it is supplying to Ukraine.

Perennial Autonomy, a company founded by Eric Schmidt, former CEO of Google, has developed two drones that have put Russian forces on the back foot — the Merops Surveyor interceptor drone and the Hornet kamikaze drone. The Surveyor interceptor uses AI to bring down Russian drones, while the Hornet drone uses both Starlink and AI to wreak havoc on Russia's ability to supply its troops along the front line.

Merops AS-3 Surveyor

The Merops AS-3 Surveyor is a mobile, truck-portable counter-drone system comprising:

1. Radar and electro-optical sensors for detection and tracking
2. A ground control/command station
3. Pneumatic or mobile launch platforms
4. A fleet of Surveyor interceptor drones

The fixed-wing Surveyor interceptor was first combat-tested in Ukraine around June 2024. By late 2025, it had reportedly achieved over 1,900 intercepts. In some sectors, it is claimed to have brought down roughly 40% of Russian Geran drones. Recent reports claim 4,000 successful Russian drone interceptions.

The Surveyor is an effective interceptor on account of its greater AI-based autonomy, speed, and jam resistance. Following launch, the drone is cued and initially guided using the sensors of the Merops system. For terminal guidance, it uses onboard IR and RF sensors, as well as AI-based machine vision. It can home in on targets even when SATNAV and communication signals are jammed.

AI-based machine vision and the ability to fuse inputs from IR and RF sensors are key to the success of the Surveyor.

With its maximum speed of 280 km/h, the drone outpaces Russian Gerans.

Hornet Strike Drone

The Hornet drone can be credited with bringing the Russian offensive in Donbas to a crawl along the line of contact, and even to a complete halt in some sectors. Ukraine is also using the drone to strangulate Russia's ability to supply Crimea.

As with the Surveyor, the Hornet's success can largely be attributed to its AI-powered ability to operate effectively in the absence of SATNAV and communications.

We covered the capabilities of the Merops Surveyor and Hornet drones in an earlier post.

AI-Based Machine Vision

So far, SATNAV has been the gold standard in the precision guidance of drones, missiles, and rockets. AI-powered machine-vision-based navigation outperforms SATNAV in accuracy. More importantly, it is completely immune to electronic warfare.

However, machine vision can be spoofed — for example, by using paint schemes that make optical recognition challenging.

With increased onboard processing power, it will become difficult, perhaps impossible, to spoof AI-powered machine vision.

Global Reach

The ability to control drones and missiles capable of precision guidance globally requires a Starlink-like network. Currently, there is no alternative to Starlink.

Outplaying Emerging Powers

In the days ahead, many nations, including India, will build weapons with AI-powered machine vision. However, doing so without acquiring matching semiconductor fabrication and design capability would not allow them to exercise sovereignty over their own weapons.

Semiconductor fabrication and design technologies are likely to be tightly controlled in order to prevent challenges to US military dominance.

As an analogy, a nation with nuclear weapons technology does not share it with a nation that lacks the technology. Indeed, nations that possess nuclear weapons do their best to prevent the "have-nots" from acquiring weapons-grade fissile material.

The ability to manufacture fissile material, a key enabling technology for nuclear weapons, is tightly controlled.

Similarly, robust space-launch capability, as well as the semiconductor fabrication and design capability needed to deploy a Starlink analogue or facilitate advanced machine vision, will be tightly controlled.

The technological barriers to acquiring these capabilities are formidable. The hurdles span multiple years — perhaps multiple decades — and are rooted in physics, engineering complexity, supply chains, and capital intensity.

A low-latency global or regional broadband constellation requires thousands of satellites (Starlink has over 10,000) in low Earth orbit (LEO, ~550 km altitude), inter-satellite laser links, high-volume satellite manufacturing, and millions of user terminals with electronically steered phased-array antennas.

The number of satellites required can vary based on the architecture of the network and its intended extent of coverage. However, a true Starlink analogue would require the development of a reusable launcher.

China, the EU, and Russia have all embarked on deploying Starlink analogues, but all three have so far made limited progress. Countries like India are unlikely to be in a position to acquire such a capability over the next decade.

User Terminals

Low-latency networks use terminals featuring custom ASICs and advanced RF front-end modules (e.g., BiCMOS technology) for phased-array antennas that track fast-moving LEO satellites.

Starlink has already deployed millions of such terminals. STMicroelectronics has shipped over 5 billion RF chips for the terminals, with daily rates exceeding 5 million.

Network satellites use radiation-hardened electronics, onboard processors, and laser comms chips that require specialized semiconductor fabs, which in turn require decades of ecosystem investment.

AI-Powered Machine Vision

Effective (high-accuracy, low-latency) machine vision in drones and cruise missiles requires real-time object detection, tracking, terrain classification, sensor fusion, and autonomous navigation capability under severe size, weight, power, and cost constraints, harsh operating environments, and contested electromagnetic conditions.

Hardware-wise, machine vision relies on high-performance AI accelerators (NPUs, custom ASICs, or optimized GPUs/FPGAs) that must deliver tens to hundreds of TOPS (trillions of operations per second) for neural networks such as CNNs or lightweight transformers (e.g., YOLO variants).

Such hardware would require leading-edge nodes (7 nm, 5 nm, 3 nm, or below) for the density, speed, and energy efficiency needed to run complex models onboard without excessive power draw.

Only a handful of fabs worldwide — primarily TSMC in Taiwan, with limited capacity from Samsung and Intel — can produce these at scale and yield.

The US itself faces geopolitical and supply-chain vulnerabilities. However, it is likely working on a plan to eventually eliminate them.

Conclusion

As things stand, the US appears uniquely positioned in combining reusable launch capability with access to semiconductor fabrication and design ecosystems that facilitate global reach and high-precision strikes by drones and cruise missiles.

In the discussion above, we confined ourselves largely to drones and cruise missiles. AI and secure global communication have applications in other weapon systems as well — space-based weapons, for example.

It is time for India to take a hard look at its quest for self-reliance in weapon manufacturing. Hopefully, we are not focusing on acquiring sunset technologies, and our efforts to acquire semiconductor fabrication and design technologies will be pursued vigorously.

Why Future Agniveers Could Well Be Disposable Humanoid Robots

Sanket Pathak, CEO Foundation Future with Phantom - a humanoid robot developed for combat tasks

CNBC reports that the US has started testing American humanoid combat robots in Ukraine.

According to the developer, Foundation Future Industries, the deployment is the first known deployment of humanoid robots in a combat zone in history.

Foundation Future is a San Francisco-based startup founded in April 2024. It develops general-purpose humanoid robots aimed at addressing labour shortages in manufacturing, logistics, domestic settings, and defence applications.

The company's flagship product — the Phantom — is roughly 5'9"–5'11" tall and weighs 176–180 lb. It is designed to lift approximately 90 lb and perform physical tasks in complex or high-risk environments. The robot has five-fingered hands, camera-first vision, and an LLM-driven autonomy system that supports both independent operation and supervised teleoperation.

Foundation Future has secured approximately $24 million in Pentagon research contracts (U.S. Army, Navy, and Air Force) for testing the humanoids.

Two Phantom units were deployed to Ukraine in February 2026 for frontline logistics and reconnaissance. Eventually, the humanoid could begin wielding weapons.

Foundation Future aims to send an upgraded humanoid variant — Phantom 2 — to Ukraine this year.

Company CEO Sanket Pathak claims that the upgraded humanoid will have "superhuman abilities" and double the payload capacity.

Pathak previously founded and led Synapse Financial Technologies, a banking-as-a-service (BaaS) fintech platform. The company filed for Chapter 11 bankruptcy in April 2024 following a disruptive collapse of its service and is reportedly under criminal investigation by the Justice Department. Eric Trump, the second son of President Donald Trump, invested in the company and, in March 2026, joined it as chief strategy adviser.

Eric's involvment in the compnay and the Pentagon contract have been frowned upon by Democrats.

In lighter vein, Foundation Future could well collapse as did Synapse. Or, it could one day tie-up with the Adani Defence to manufacture disposable Agniveers for the Indian Armed Forces!