[against Class I FPV drones]
The only sustainable response to fiber-optic FPV.
THE RULES HAVE CHANGED
In three years, FPV kamikaze drones have become the dominant tactical weapon in asymmetric conflicts.
Ukraine is not the only theater.
Iran-backed and non-state actors are deploying FPV kamikaze drones worldwide.
Established air defense fails — repeatedly.
FPV kamikaze drone strikes US military aircraft.
An Iran-backed FPV kamikaze drone struck a US military aircraft at the base near Baghdad airport for the first time — a helicopter and a Sentinel radar hit. Multiple follow-up strikes within days, no conventional air defense could intercept the Class-I platforms.
First IDF soldier killed by FPV drone.
A Hezbollah FPV kamikaze drone with fiber-optic control — structurally immune to RF jamming — struck an Israeli armored unit. 90 fiber-optic FPVs in three months against IDF positions, Iron Dome and Electronic Warfare ineffective.
First documented FPV strike on an Iron Dome position.
A Hezbollah fiber-optic FPV drone strikes an Iron Dome position in northern Israel. The IDF did not deny the strike footage published by Hezbollah; the Israeli military responded by deploying thousands of meters of fishing nets across southern Lebanon. When the world's most advanced air defense system can no longer protect itself, the asymmetry is undeniable.
How the Herakles interceptor neutralizes the threat.
Pick an attack pattern and the asset to defend — the simulation shows each run from detection to hit.
WHY EXISTING SYSTEMS FAIL
For every category we name the competing systems — with vendor claims, sources and dates.
Conventional jammers reach their limits — modern FPVs hop frequencies, and fiber-optic FPVs are increasingly adopted and structurally immune to RF jamming. And a second development is spreading: more and more FPV drones find their target in the final approach on their own, with AI on board. Even cutting the radio link no longer stops them — they complete the attack without one. That removes the radio link as a path to the target: what is left has to reach the drone itself. High-power microwave can do that, but only from a fixed position and across an area. Against a single manoeuvring target at close range, what remains is the physical hit.
Competitors and sourcesNATO ACT · DroneShield · Auterion · The Fourth Law · Epirus
NATO ACT · Innovation Challenge, ACT-SACT-25-48
FindingOn fibre-optic controlled FPV drones: “EW counter-UAS systems are ineffective against this type of drone.” And: “The combination of high maneuverability, jamming immunity, and low visual and radar signature makes these drones particularly dangerous to frontline troops.”
Sources:Unmanned Airspace, 2025-05-03 (zu NATO ACT, RFIP ACT-SACT-25-48)Defense News, 2025-06-12
DroneShield · RF detection and jamming
The maker’s own statementSix months after that finding, the company’s Strategic Adviser Gus McLachlan states: “We still remain of the view that a large portion of the threat will be defeated by interrupting that RF communication between the device and the operator.”
Not statedA demonstration against a fibre-optic guided drone is not named in the published material.
Auterion · Skynode S, terminal guidance for 50,000 FPV drones
The maker’s own statement“Operators gain precision against moving and evasive targets in contested electronic environments, including conditions where GPS and the radio link are degraded or denied.” The radio link is no longer a precondition on the attacking side either.
Sources:Auterion, 2026-07-13
The Fourth Law · TFL-1 — retrofit terminal guidance
FindingThe module “provides autonomous terminal guidance for the final 400 to 500 meters” of an FPV flight and is “built to endure heavy jamming by ignoring the radio horizon”. The price: “Each TFL-1 costs under $100, keeping the upgrade well below 10 percent of a typical FPV drone.”
Shown publicly“TFL-1 has been certified for military use by the country’s defense ministry and is now in service with roughly 20 brigades.” Not a prototype and not a concept, but certified and in service.
Sources:The Defense Post, 2025-11-19
Epirus · Leonidas — the counter-evidence we name ourselves
The maker’s own statement“The event marks the first known instance of electromagnetic interference being weaponized to defeat a fiber-optic guided drone.” Fibre optics are therefore not immune to every electronic effect.
Not statedWhat the fibre-optic guided drone was doing during the engagement is not stated in the release. Nor is whether it was cued externally.
Sources:Epirus, 2026-01-13
Our readingWhat is dead is the radio link as a path to the target, not every electronic effect. High-power microwave hits the electronics rather than the link — but it is position- or vehicle-bound and works across an area, not as a point intercept of a single manoeuvring target.
As of 19 Aug 2026. Quotations are verbatim from the source named. Missing information is marked as missing, not interpreted.
Conventional air defense is designed for larger platforms. FPV kamikaze drones are too small, too fast, too agile for targeted hits. Plus: due to their sheer size these systems are immediately spotted, draw fire and become targets for drone salvos that eventually overpower them. And the arithmetic does not work out: one gun per position faces a threat that arrives in numbers.
Competitors and sourcesAllen Control Systems · ZeroMark
Allen Control Systems · Bullfrog
The maker’s own statement“Bullfrog™ is an autonomous gun turret that detects, identifies and neutralizes adversarial UAVs.” Sensors, range and target speeds are not stated on the maker’s page.
Not statedOf the Project Convergence Capstone 5 demonstration it is recorded that it was “confirming kills on all seven drone targets within range during a breach scenario”. What those seven targets flew — profile, speed, evasive behaviour — appears in none of the published reports.
ZeroMark · Fire Control System
FindingThe War Zone in October 2024: “It uses an array of sensors, including electro-optical cameras and LIDAR, coupled with machine vision and advanced software algorithms, to acquire targets.”
Not statedOn the flight behaviour of the targets in the 3rd Marine Division live-fire demonstration, the reporting says nothing.
Sources:The War Zone, 2024-10-24
The standard the trade press itself applies
FindingDroneXL in July 2026, on another maker’s intercept demonstration: “publish the same demonstration against truly fast aircraft. Show me the intercept timeline against a drone flying an attack profile, not a camera platform holding a line.” And: “It means the public evidence shows a kill against the easier end of the problem.”
Sources:DroneXL, 2026-07-18
Our readingWhat is in question is not the capability but the proof and the arithmetic. A turret of this class weighs close to two tonnes, needs external radar cueing, and faces a threat that arrives in numbers.
As of 19 Aug 2026. Quotations are verbatim from the source named. Missing information is marked as missing, not interpreted.
Designed against Class II/III long-range drones — proven against loitering munitions and larger platforms, as an alternative to a guided missile. Against Class I FPV threats at close range, the geometry breaks down: minimum airspeed and turn radius are too large for agile quad targets that exploit proximity and surprise. They also depend on external cueing: without an upstream radar such an interceptor never finds its target — and that radar is itself an emitting, locatable position. Plus logistical overhead: launch rails, a radar unit and pre-flight assembly make them impractical for mobile frontline force protection.
Competitors and sourcesQuantum Systems / WIY Drones · Quantum Systems · TYTAN Technologies · MARSS · Raytheon · Rafael / SpearUAV · Anduril
Quantum Systems / WIY Drones · STRILA
The maker’s own statement“The STRILA interceptor UAV is designed to actively counter enemy wing-type UAVs, protecting positions, units, and critical infrastructure from aerial reconnaissance and attacks.” On cueing: “Radar system and radar station (RLS) integration … enables automatic cueing of interceptors toward detected targets.”
Not statedThe brochure names fixed-wing aircraft and reconnaissance as the target class. Small, manoeuvring quadcopters do not appear in the document.
Quantum Systems · Skyron 400
The maker’s own statement“Skyron 400 is a high-speed, cost-efficient interceptor UAV purpose-built to counter long-range loitering munitions like Shahed-136 drones.” On cueing: “Target detection and cueing is handled through external radar sensors integrated into the MOSAIC command-and-control workflow. The standard configuration includes a fully validated Echodyne Shield radar.”
Not statedThe page claims to cover “the full spectrum of drone threats” — yet the threats it names throughout are the Shahed-136, loitering munitions and medium-altitude ISR drones. No FPV or quadcopter target class appears. The maker gives the radar detection range as up to 10 km and the take-off weight as 4.5 kg.
Sources:Quantum Systems, Skyron 400
TYTAN Technologies · TI-1 METIS and TI-2 EOS
The maker’s own statementFor both interceptors the maker’s page names “Ground Radar & Autonomous Navigation” as the guidance principle; the TI-2 EOS is, by the maker’s own account, built for NATO Class II threats.
Not statedThe maker’s page, the system animation and the reporting all describe a layered defence against Shahed-type drones. FPV targets do not appear in any of it.
Sources:TYTAN Technologies
MARSS · Interceptor-MR
The maker’s own statement“The Interceptor is capable of defeating fast, high manoeuvring targets.” On cueing: “housed in a vertical smart launcher connected to the NiDAR sensor infrastructure. Once a threat is detected, the operator deploys the Interceptor from the launcher.”
Not statedThe product page carries no demonstration evidence for the manoeuvring-target capability it claims.
Sources:MARSS, Interceptor-MR
Raytheon · Coyote
The maker’s own statement“Raytheon’s Coyote family of effectors pairs with the Ku-band Radio Frequency Sensor, or KuRFS, to deliver advanced detect-and-defeat capabilities against unmanned aircraft systems.”
Not statedQuadcopters and FPV-class targets are not named on the product page.
Sources:RTX / Raytheon, Coyote
Rafael / SpearUAV · Iron Wasp
FindingEDR Magazine headlined the Eurosatory 2026 presentation “Rafael and SpearUAV virtually launch the Iron Wasp C-UAS effector” and writes: “Although not present in mock-up form, Israel unveiled the Iron Wasp”. France had imposed restrictions on Israeli stands. A live demonstration for international customers was planned at that point, not held.
Our own reviewThe announcement video on Rafael’s own channel, published on 16 June 2026, shows an animated concept throughout. It contains no live footage of an intercept.
Sources:EDR Magazine, 2026-06-25Israel Defense, 2026-06-22„Just Launched: IRON WASP C-UAS“, RAFAEL Advanced Defense Systems, 2026-06-16
Anduril · Anvil
The maker’s own statementAnvil “links with the artificial intelligence-empowered Lattice command-and-control software” and is aimed at Group 1 and Group 2 systems. The external cueing is advertised, not concealed.
Our own reviewIn the system video on Anduril Industries’ own channel, published in January 2022, the interceptor hits a drone of the DJI Phantom class holding station motionless at 1:40. The same sequence runs as a loop on the maker’s website. In the text reporting, the flight behaviour of the targets remains undisclosed to this day.
Sources:C4ISRNET, 2023-10-05„Anduril Lattice Counter Drone System“, Anduril Industries, 2022-01-24, 1:40
Our readingEvery one of these chains begins at a sensor that is not on board. And wherever the target class is named at all, it is the Shahed and reconnaissance domain — not the small, manoeuvring quadcopter at close range.
As of 19 Aug 2026. Quotations are verbatim from the source named. Missing information is marked as missing, not interpreted.
The systems common across the industry sit on large, sluggish carrier platforms and are built against the previous threat generation: slow, often hovering Mavic-class camera drones. They need elaborate sensors — usually LiDAR — and an approach phase that a highly agile, maneuvering FPV never grants. Against a target that changes its path by the second, the capture comes too late. And every intercept attempt ties up an entire platform — the threat comes in numbers, the countermeasure does not.
Competitors and sourcesARGUS Interception · Fortem Technologies · Alpine Eagle
ARGUS Interception · A1-Falke
The maker’s own statementThe stated purpose is the “controlled interception and secure recovery of uncooperative drones”; the system “brings them to the ground intact, enabling forensic evaluation and preservation of evidence”. The effector carries radar, LiDAR and a depth-sensing camera.
Our own reviewIn a public press demonstration filmed by euronews in June 2026, a near-motionless Holybro X500 V2 is captured — an open development quadcopter, not a threat representative airframe. The same scene runs as a loop on the maker’s landing page.
Sources:ARGUS Interceptioneuronews (deutsch), „Mit Radar und Fangnetz“, 2026-06-01, 0:23
Fortem Technologies · DroneHunter F700
The maker’s own statement“Utilizing its onboard TrueView® R20 radar, the F700 detects, tracks, and locks onto its target with conviction.” Capture is by net: a tether net against Group-1 drones, and “DrogueNet™, a device that disables large Group-2 drones”.
Not statedTarget speeds are not stated on the product page. For the more than 4,500 captures claimed, not one drone type is named.
Alpine Eagle · Sentinel
Finding“Sentinel combines Alpine Eagle’s airborne radar and sensor network with a software-defined defence architecture to detect and track hostile drones across wide areas and defeat them with onboard airborne interceptors.”
Not statedTarget classes, FPV suitability and the flight behaviour of targets in the trials are not stated in the published material.
Sources:European Security & Defence, 2026-03-19Alpine Eagle, Sentinel
Our readingA capture needs an approach phase and a target that allows one. Where the stated design intent is intact recovery for evidence, it is not the interception of an incoming weapon.
As of 19 Aug 2026. Quotations are verbatim from the source named. Missing information is marked as missing, not interpreted.
Promising complementary capabilities, but vehicle- or platform-bound and power-intensive. Lasers are additionally constrained by atmospheric conditions (fog, rain, dust) and require dwell time per target: for that duration the beam has to stay on the exact same spot of the target — barely achievable against a fast-maneuvering FPV. Neither is an infantry-portable answer for mobile frontline force protection. Plus: due to their size and platform binding they are immediately spotted — becoming targets for drone salvos that eventually overpower them. And what each position demands in power and platform cannot be multiplied at the pace the threat multiplies.
Competitors and sourcesINLEAP Photonics · Epirus
INLEAP Photonics · FASTLIGHT SHIELD
The maker’s own statement“The laser beam can be steered up to 2500 times faster than with conventional beam steering systems.” Target class, engagement time and cueing architecture are not stated on the maker’s page.
Our own reviewTwo television documentaries show the same footage — ntv in July 2026 and RTL Nord in August 2026: a DJI Mini 3 hovering motionless roughly three metres above the ground is defeated after several seconds of continuous irradiation of the same spot. The effector sits on a trailer.
Sources:INLEAP Photonicsntv nachrichten, 2026-07-14, 0:27RTL Nord, 2026-08-03, 3:13
Epirus · Leonidas
The maker’s own statement61 of 61 drones in a live-fire demonstration, described by the maker as flying “operationally relevant flight scenarios”.
Not statedWhether the targets evaded or manoeuvred is not stated in the release. The effect is an area effect against swarms, not a point intercept of a single target.
Sources:Epirus, 2025-09-10
What the physics demands
FindingPeer-reviewed: “The laser technique is affected by adverse weather conditions, such as fog, rain, and cloud.” And: “Drones are usually fast and agile, so destroying them should require a short time and high energy.” The effect is proportional to power and to time on target.
Sources:Chaari & Al-Maadeed, Security and Defence Quarterly 32(5), 2020
Our readingDwell time is not a side condition, it is the arithmetic itself: the beam has to stay on the same spot of the target for seconds. What the published demonstrations put against it is a target that holds still.
As of 19 Aug 2026. Quotations are verbatim from the source named. Missing information is marked as missing, not interpreted.
ARCHITECTURE PRINCIPLE
THE ANSWER: DRONE VS. DRONE.
Radically reduced by design.
An autonomous drone that intercepts FPV drones. It launches vertically from a standing start — no vehicle, no ramp, no command post — and works by pure collision, without a warhead. Designed against a target that evades, can't be jammed, and comes in numbers:
Built like its target
An FPV drone dodges, changes direction by the second and flies low. Catching it takes the same agility. So we use the airframe the adversary uses — except ours carries no warhead and flies autonomously. Less mass, more margin: our interceptor dictates the maneuver, not the attacker.
Works where jammers stop
The standard answer to a drone is to break its radio link. Fiber optics and onboard AI in the final approach have retired that answer. Our interceptor never goes after the link at all: it sees its target and flies at it — no radio, no GPS, no one else pointing the way.
Affordable in numbers
This threat arrives in vast numbers. A defense that costs many times the attack does not add up. Because our interceptor does without expensive sensing and without explosive payload, it sits in the same price class as what it takes down.
COMPLIANCE & SYSTEM FRAME
Deployment Contexts
Bundeswehr & NATO Forces
Infantry and convoy protection at tactical close range. Vertical Takeoff, infantry-deployable, kinetic effector under Operator-Authorization.
EU Allied Forces
Counter-UAS layer for armed forces facing active Class I FPV threats. Established defense validation channels. ITAR-free; dual-use across defense and critical infrastructure.
Critical Infrastructure (DE/EU)
Airports, energy infrastructure, ports, major events. Detection-first mode; kinetic deployment exclusively via an authorized state operator.
FROM SIM TO INTERCEPT.
[ SIM ]
In-house simulation environment with full flight dynamics, calibrated sensor-noise models and parameterized threat profiles. Batch-evaluation across thousands of engagement scenarios against evasive, accelerating and sensor-noisy targets. Identical code path between simulation and flight hardware closes the classic sim-to-real gap.
[ BENCH ]
Standardized bench routines for every hardware iteration: powertrain characterization, sensor calibration, failsafe verification and authorization logic under realistic load profiles. Automated regression tests safeguard function and safety before any field deployment.
[ REAL-FLIGHT ]
Controlled flight campaigns on secured test ranges. Fully autonomous intercept of an inbound, actively maneuvering FPV threat under operator authorization — acquisition, tracking and intercept course computed on board, without GPS, without a rangefinder and without a radio link. The target: a fiber-optic-controlled 10-inch FPV, manually piloted and evading laterally; engagement from approx. 70 m at approx. 25 m/s closing speed. Reproduced multiple times, documented on video; raw footage and flight data under NDA. Every real flight feeds empirical data back into sim and bench.