Quick Reference
The six facts readers search for most, answered in one line each. Details and sources are in the sections below.
YFQ-42A (General Atomics, "Dark Merlin") and YFQ-44A (Anduril, "Fury") are USAF Collaborative Combat Aircraft Increment 1. YFQ-48A ("Talon Blue") is real too — it belongs to Northrop Grumman, targets Increment 2, and is not yet flying.
Digital RF Memory: a jammer digitally samples an incoming radar pulse, then retransmits a delayed or altered copy to paint false targets the source radar can't easily distinguish from real ones.
Runs on a MD‑550 piston engine (~50 hp), not a rocket. "Specific impulse" is a rocket-motor metric — it doesn't apply to the cruise engine, which is why no legitimate Isp figure for it exists.
A gyro-stabilized turret that fuses a daylight camera, thermal/IR imager, and often a laser rangefinder or designator on one gimbal mount — used for tracking, surveillance, and targeting.
Almost certainly a mis-transcription of anti-heeling system — a Kongsberg Maritime ballast-stabilization product for ships, unrelated to air defense. Kongsberg's actual air-defense flagship is NASAMS.
Rafael has never published an official per-interceptor price. Press estimates range roughly $40,000–$150,000+ depending on batch and surge conditions — treat any single number as an estimate.
The Counter-UAS Kill Chain
Every counter-drone system slots into one of four stages. No single vendor covers all four end-to-end — most programs stitch sensors and effectors from different companies together.
Detect
Passive RF direction-finding, radar, and acoustic sensors pick up a drone's presence.
- DroneShield RfPatrol
- Hidden Level passive RF network
- Dedrone RF-360 / RF-160
- GA-ASI Lynx SAR/GMTI radar
Track & Fuse
RF, radar, and EO/IR tracks are correlated into one shared picture.
- Anduril Lattice
- Fortem SkyDome Manager
- Kongsberg / APS PROTECTOR CUAS
- Dedrone sensor-fusion C2
Classify
AI/ML models separate hostile drones from birds, aircraft, and clutter.
- Fortem TrueView microdoppler AI
- Saab Giraffe 1X "Drone Tracker" mode
- Robin Radar IRIS
- Lattice AI-scored threat confidence
Engage
A kinetic, electronic, or directed-energy effector defeats the confirmed threat.
- Anduril Anvil (kinetic ram)
- Fortem DroneHunter (net capture)
- DroneShield DroneGun (RF jam)
- Epirus Leonidas (HPM) / Lockheed HELIOS (laser)
Detection, Tracking & Sensor Fusion
"CUAS sensor fusion" means combining RF direction-finding, radar tracking, and EO/IR classification into one AI-scored track, often with automated cross-cueing between sensors. These vendors build that layer.
RfPatrol Mk2 & DroneSentry-X Mk2
Passive, non-RF-emitting wearable SDR-based drone detector plus a software-defined fixed/mobile detection-and-disruption system. Deployed in 40+ countries.
- Battery life Up to 8 hours (RfPatrol Mk2)
- Operating range −20°C to +60°C
- FY2025 revenue >AUD 260 million (company-reported)
Source: DroneShield product pages; company FY2025 results.
RF-360 / RF-160 sensor fusion
Acquired by Axon in October 2024. Fuses RF sensors, radar, and PTZ cameras — "one sensor identifies, the other verifies." Detection/classification claimed across 150+ drone manufacturers.
- RF-360 range ~2 km LOS, up to 5 km ideal
- RF-160 range ~1.6 km avg, 5 km max
- Setup time ~20 min (mobile units)
Source: dedrone.com; Axon acquisition announcement, Oct 2024.
Passive RF sensing network
No active emissions; detects, tracks, and identifies via passive RF/passive-radar. Company states TRL-9 (fully operational/deployed).
- DoD funding $10M APFIT award for Tactical Passive Radar
- Validation Two-year USAF evaluation, live Group 1 UAS tracking
Source: PR Newswire, Hidden Level press releases.
TrueView R30 / R20 radar
Compact 3D AESA radar using NVIDIA GPU-powered AI microdoppler classification (trained on 10+ years of ML data) to filter birds from drones. Four units combine for 360° "SkyDome" coverage.
- Weight <7 kg
- Power draw 38 W (R20)
- Detection range >1,500 m (R30)
Source: fortemtech.com/products/trueview-r30/; DroneLife, Mar 2026 (Lockheed Sanctum integration).
PROTECTOR CUAS
Remote weapon station integrating radar, jammers, guns/cannons, and hunter-UAVs, networked to share a radar picture across units; APS's C2 fuses radar, EO, and acoustic sensors.
- Poland order NOK 16B (~$1.5B) for 18 CUAS batteries, signed 30 Jan 2026
- Program SAN (Systemy AntyDronowe), with PGZ; deliveries expected 2027
- Effectors 35 mm / 30 mm / 12.7 mm guns, missiles, interceptor drones
Source: kongsberg.com press release (NOK 16B); Army Recognition and Defense News report the same award as ~$1.5B for 18 batteries — one contract in two currencies, not two deals. Wider PLN 15B (~$4.2B) figures cover Poland's full SAN program, not Kongsberg's share.
Giraffe 1X
Compact 3D AESA radar marketed as a C-UAS "Drone Tracker" with a software-updatable threat library. In service/on order with the US, Sweden, Latvia, Denmark, Belgium.
- Weight <150 kg total (~100 kg topside)
- Detects RCS <0.01 m² out to 75 km (small UAV ~4 km)
- Refresh ~1 sec full 360°
- US Army order $46M (2025)
Source: saab.com/products/giraffe-1x; Army Recognition, 2025.
IRIS radar
Counter-drone/bird-discrimination radar distinguishing small, high-speed, low-signature drones from bird clutter. Deployed for 2026 FIFA World Cup security and tested in Ukraine.
- Weight 29 kg
- Range switchable 5 km / 12 km (Long-Range Mode, Sept 2025)
Source: The Defense Post, Sept 22, 2025; robinradar.com.
Lattice-fused sensor towers (Sentry)
Solar-powered towers combining wide-area radar, EO/IR turret, and RF detection with edge AI classification. One vendor among many in this layer — described here strictly on the same factual basis as the others.
- Extended Range Sentry Tower Detection to 7.5 mi (5+ mi autonomous), 80 ft mast
- Maritime Sentry Tower ~9 mi detection radius for small boats
- Contract $363M CBP award for XR Sentry Towers
Source: anduril.com; Defense News, Oct 2022; Tectonic Defense contract coverage.
Kinetic Drone Interceptors
Physical-defeat systems: ramming, netting, or explosive interceptors that destroy or capture a hostile drone rather than jam or dazzle it.
Anvil / Anvil-M
Autonomous quadcopter interceptor engaging Group 1–2 UAS by physical collision, cued by Lattice sensor fusion with human authorization for final engagement. Anvil-M (Oct 2023) adds a fire-control module and explosive payload for harder/faster targets.
- MADIS contract $200M, 5-year, US Marine Corps (Nov 2024)
- Follow-on $642M Marine Corps counter-drone contract (Mar 2025)
Source: anduril.com/anvil; The Defense Post, Nov 2024 & Mar 2025. Precise range and unit cost are not publicly disclosed.
DroneHunter F700 / 5.0
Autonomous, radar-guided interceptor drone that captures targets with expanding tether nets or a DrogueNet® parachute-net combo, rather than destroying them.
- Field record "More than 4,500 drone captures" (company-stated)
- Reload Under 3 minutes for back-to-back sorties
Source: fortemtech.com/products/dronehunter-f700/; press release, Jan 20, 2026.
Interceptor-MR / -SR
Kinetic hard-kill drone that rams hostile UAS with a reinforced nose cone rather than an explosive warhead. The tail-sitting VTOL Interceptor-MR defeats Category I & II drones head-on.
- Weight 8 kg (Interceptor-MR)
- Top speed ~155 kt (288 km/h)
- Engagement Up to 5 km+, above 2 km altitude
Source: MARSS Interceptor product documentation; Aerotime interview.
Drone Dome (hard-kill configuration)
Combined kinetic/electronic system offering both a laser hard-kill effector ("Drone Dome-L") and VHF/UHF/GNSS jamming soft-kill, cued by a 360° radar and RF sensor under an AI-enabled C4I layer.
- Detection Beyond 3.5 km
- Status US DoD-certified (Nov 2022)
- Notable use 2018 G20 (Argentina), 2018 Youth Olympics
Source: rafael.co.il; C4ISRNET, Nov 28, 2022.
Could not verify: a Chinese kinetic ram/net-capture interceptor drone directly comparable to Anvil, DroneHunter, or MARSS Interceptor. Chinese systems found in this research (AVIC Light Arrow/Sky Shield, CASIC LW-30/LW-60) are laser or jamming systems, not kinetic interceptors.
Electronic Warfare & Directed Energy
How FPGA/SDR-based radios turn into missile jammers, what DRFM actually does, and the directed-energy systems now defeating drone swarms.
FPGA + SDR + ESM + ECM: how a missile jammer actually works
ESM (Electronic Support Measures) is the passive receive chain: it detects, characterizes, and geolocates hostile RF/radar emissions without transmitting. ECM (Electronic Countermeasures) is the active transmit chain: it jams or deceives those same threats. SDR (software-defined radio) is the wideband digital RF front end that lets one hardware chassis be reconfigured in software to do ESM one moment and ECM the next. The FPGA inside that SDR is the real-time processing core — it digitizes, buffers, and generates waveforms fast enough (nanosecond-to-microsecond latency) to keep pace with an inbound missile seeker, which a general-purpose CPU cannot do.
Vendors building FPGA/SDR-based ESM/ECM for missile-jamming and self-protection: Leonardo (Miysis DIRCM against IR-homing missiles/MANPADS; BriteCloud DRFM decoy), L3Harris (ASPEN — a software-configurable signal-processing platform across its EW/CDL lines), Leonardo DRS (DAIRCM helicopter self-protection suite), Elbit Systems (Light SPEAR / Spectrolite, built on multiple parallel DRFM channels, full-band low-band–18 GHz coverage), BAE Systems (PRISM ESM family; Nulka active missile decoy on 150+ USN/RAN/RCN warships), and Northrop Grumman (AN/ALQ-249 Next Generation Jammer Mid-Band, IOC 2021 on EA-18G Growler).
Source: Leonardo, L3Harris, Leonardo DRS, Elbit, BAE, Northrop Grumman product pages; MDPI Applied Sciences 10(12):4123 (FPGA DRFM design).
DRFM (Digital RF Memory) and electronic countermeasure types
DRFM digitally samples an incoming radar/RF waveform, stores it coherently in memory, then retransmits a delayed and/or modified copy — creating false targets or masking real ones in a way the emitting radar cannot easily distinguish from a genuine return. It is the dominant modern deception technique because it can replicate coherent, convincing false targets.
The broader ECM toolkit: noise/barrage jamming (broadband RF power overwhelms a receiver's signal-to-noise ratio), deception jamming (range-gate or velocity-gate pull-off feeds false position/velocity data), chaff (cut metallic strips creating false radar returns, tuned by length to different threat frequencies), and DRFM repeaters (above).
Named DRFM systems: Leonardo BriteCloud (expendable, self-contained DRFM jammer/decoy for fast jets), Elbit Light SPEAR/Spectrolite (multi-channel DRFM ESM/ECM). Northrop Grumman's ALQ-249 uses digital-technique generation consistent with DRFM patent literature, though Northrop's own materials don't label it "DRFM" explicitly.
Source: Wikipedia (DRFM); radartutorial.eu; Association of Old Crows; leonardo.com; elbitsystems.com.
Leonidas (high-power microwave)
Counter-swarm HPM system under the U.S. Army's IFPC-HPM program. Initial award $66.1M (Jan 2023); four Gen-I prototypes delivered by March 2024, deployed to CENTCOM by early 2025.
- Gen II award $43.55M (Jul 17, 2025) for two systems
- Gen II claims "More than double" range, "~30% more power" vendor claim
Source: army-technology.com; epirusinc.com. Power output in kW/MW is not independently verified — MIT Technology Review (May 2025) reports the company's own near-term goal is only "tens of kilowatts."
UNVERIFIED POWER FIGUREHELIOS laser
High Energy Laser with Integrated Optical-dazzler and Surveillance. 60 kW-class, expandable 60–120 kW via modular fiber-laser combination, plus a counter-UAS dazzler and ISR sensor suite.
- Fielded USS Preble (DDG-88), Aug 2022 — first tactical laser on an existing warship
- Test record Downed 4 drones in Oct 2025 testing
Source: lockheedmartin.com; TWZ, Oct 2025; Naval News, Feb 2025.
HELWS (High Energy Laser Weapon System)
Product family spanning a 10 kW-class palletized "H4," a 15 kW-class HELWS variant, and a 50 kW-class laser demonstrated on a Stryker at White Sands.
- First overseas USAF deployment 2019
- Claimed record 40,000+ test hours, 400+ downed targets vendor claim, not audited
Source: rtx.com, Feb 2026; Breaking Defense, Jul 2025.
UNAUDITED CLAIMIron Beam laser
100 kW-class laser weapon system. First operational unit delivered to the IDF Dec 28, 2025. Reported first combat interception of projectiles from Lebanon, Mar 2, 2026.
- Power class 100 kW (consistent across vendor + independent defense press)
Source: rafael.co.il; National Security Journal, Dec 2025; ASD News, Dec 28, 2025.
Pulsar (RF sensing + jamming)
Passive ESM detection plus active RF jamming node, software-defined and networked — not a directed-energy weapon. One vendor among several in this category.
- Latest variant ~2 km engagement range, IP65, −30°C to +65°C, <4 kW draw (World Defense Show 2026)
Source: armyrecognition.com, WDS 2026 coverage; anduril.com/pulsar.
EO/IR Gimbals & Radar Payloads
An EO/IR gimbal is a gyro-stabilized, multi-sensor turret — daylight camera plus thermal/IR imager, often with a laser rangefinder or designator — mounted on aircraft, drones, vehicles, or vessels.
Star SAFIRE 380-HD
Multi-sensor turret with a cooled InSb MWIR imager and up to 120x optical magnification.
- Resolution 640×512 std, 1280×720 optional
- Weight <99 lb (<45 kg)
- Power 280 W (350 W max)
Source: defense.flir.com datasheet.
MX-15 / MX-10
MX-15: 4-axis stabilized, up to seven simultaneous sensor payloads. MX-10: smaller dual-sensor variant sharing MX-15/MX-20 tech.
- MX-15 ~100 lb (45 kg), 15.5″ dia
- MX-10 38 lb (17 kg), 10″ dia
- Contract $96M SOCOM IDIQ, May 2026
Source: l3harris.com; Vertical Mag, May 2026.
TASE 400 series
Compact gimbal (formerly Cloud Cap Technology) with integrated GPS/INS, 360° continuous pan, and onboard target tracking.
- Weight 8 lb (3.62 kg)
- Slew rate 150°/sec
- Field record 500,000+ operational hours (company-stated)
Source: cloudcaptech.com; collinsaerospace.com.
Euroflir 410
Hosts up to 10 sensors and up to 4 lasers across 4 spectral bands (TV, NIR, SWIR, MWIR). Variants: Surveillance, Marking, Designation.
- Laser rangefinder/designator To 20 km (1.54 μm)
- Fielded on French Navy H160 helicopters (2020–present)
Source: safran-group.com brochure; Naval News, Sept 2020.
Micro-CoMPASS (on Skylark II)
8″ gimbal with a day sensor and cooled night/thermal sensor, laser illuminator, and laser tracker, integrated for single-operator fly-by-camera control.
Source: elbitsystems.com/product/skylark-i-lex/. Precise resolution/range figures not published.
Lynx Multi-Mode Radar
Ku-band SAR/GMTI radar flown on MQ-9 Reaper (USAF, RAF, Italian AF, French AF) and Army Gray Eagle/Sky Warrior.
- SAR resolution 0.1 m spotlight, 0.3 m stripmap
- Weight <120 lb, 320 W peak output
- Range >45 km in 4mm/hr rain
Source: GA-ASI Lynx datasheet (PDF).
Multi-Mission Hemispheric Radar (MHR)
Counter-UAS / short-range air defense radar product line, alongside ground-vehicle EO/IR systems (Family of Weapons Sights, 3rd-Gen FLIR for the Army).
- Recent contracts Up to $800M total across multiple IDIQs (2025–2026)
Source: leonardodrs.com. Detailed frequency/range specs not publicly listed.
"EO/IR gimbal market" size — and why the numbers disagree
No single authoritative government or trade-association figure exists for this market — only competing commercial market-research estimates. IMARC Group puts the global EO/IR gimbal market at $957.7M in 2025, forecast to reach $1.95B by 2034. Global Growth Insights puts it at $676.48M in 2026, forecast to reach $720.18M in 2027 (~65% defense / 35% civil split, ~40% North America share). These two paid-report publishers disagree by roughly 30–40% — typical for niche defense-market forecasting, where methodology and scope vary. Treat both as directional estimates, not verified facts.
Source: imarcgroup.com/eo-ir-gimbal-market; globalgrowthinsights.com/market-reports/eo-and-ir-gimbal-market-104186.
"Iris sensor" — which one do you mean?
In a defense EO/IR context, "Iris" maps most concretely to Anduril's Iris passive IRST/targeting sensor family (announced 2024) and, secondarily, to Robin Radar's IRIS counter-drone radar (an unrelated product from a different company). Generic "iris sensor" searches without defense qualifiers mostly surface an entirely different field — biometric iris recognition (identity scanners), which has no connection to either defense product. If you landed here from a defense search, you almost certainly mean one of the two products above, not a biometric scanner.
Collaborative Combat Aircraft (CCA) & Loyal-Wingman Drones
The USAF's CCA program pairs uncrewed "loyal wingman" jets with crewed fighters. Designations have shifted more than once — here's what's actually confirmed as of mid-2026.
| Designation | Maker | Nickname | Program | Status | Published performance |
|---|---|---|---|---|---|
| YFQ-42A | General Atomics Aeronautical Systems | Dark Merlin | CCA Increment 1 | First flight Aug 27, 2025; EMD/production award Jun 17, 2026 | Speed/range/endurance not publicly released. |
| YFQ-44A | Anduril Industries | Fury | CCA Increment 1 | First flight Oct 31, 2025; EMD/production award Jun 17, 2026 | Endurance "15+ hrs," speed ~Mach 0.95 — vendor/press estimate |
| YFQ-48A | Northrop Grumman | Talon Blue | CCA Increment 2 contender | Designated Dec 22, 2025; autonomous taxi tests May 2026; first flight targeted late 2026 | Not published. Not part of Increment 1 — a separate program from YFQ-42A/44A. |
| — | Kratos Defense | XQ-58A Valkyrie | AFRL demonstrator (LCASD / LCAAT), not USAF CCA | First flight Mar 5, 2019; now also flying with USMC | Max 566 kn (1,048 km/h); cruise 476 kn (≈Mach 0.72); range ~3,000 nmi; ceiling 45,000 ft — Kratos datasheet |
| MQ-28A | Boeing Defence Australia | Ghost Bat | RAAF program — not USAF CCA | First live AIM-120 firing Dec 8, 2025; Block 3 build starts 2027 | Range >2,000 nmi; speed to Mach 0.9; ceiling >40,000 ft — Boeing/trade-press |
Program structure
- Two vendors (GA, Anduril) were both picked for CCA Increment 1 EMD + production on June 17, 2026, beating at least three unnamed competitors.
- A separate autonomy-software competition advanced Anduril, Shield AI, and Collins Aerospace to the next phase (DefenseScoop, Jun 2026).
- Combined near-term goal: 150+ CCA aircraft by end of decade (SecAF Troy Meink) — officially confirmed.
The "1,000 CCA" figure
- Originated with then-SecAF Frank Kendall, March 2023, as an explicit planning figure — roughly a 2:1 ratio of CCAs to NGAD/F-47 and F-35 airframes.
- Not a program of record. Kendall separately cited ~100 CCAs by 2029 as the near-term production goal, with up to 2,000 floated as a long-term possibility contingent on cost.
Common mix-ups
- MQ-28 Ghost Bat is not a USAF program. It's Australian/Boeing-led (RAAF), with no link to the CCA competition despite flying alongside US aircraft in joint exercises.
- "Cruise speed Mach 0.8" has no confirmed single-aircraft attribution in official sources. The closest documented figure is XQ-58A Valkyrie's Mach 0.72 cruise (Kratos datasheet) — don't treat 0.8 as a confirmed spec for any named platform.
- General Atomics' "Gambit" family (Gambit 1–6) is GA's internal product branding, not a separate program of record from YFQ-42A.
Loitering Munitions
"Kamikaze" or "suicide" drones that loiter over a search area before diving into a target. Specs below separate manufacturer datasheets from open-source/OSINT estimates — the latter dominate for adversary systems with no public spec sheet.
Shahed-136 "specific impulse" — why no such figure exists
The Shahed-136's cruise propulsion is the Mado MD-550, a reverse-engineered copy of the German Limbach L-550E: a 550cc, four-cylinder, two-stroke, air-cooled piston engine producing roughly 50 hp (37 kW) at 7,500 rpm, driving a pusher propeller. Specific impulse (Isp) is a rocket-motor metric — thrust per unit propellant weight-flow, in seconds — and it simply does not apply to a reciprocating propeller engine, which is instead rated by horsepower and fuel consumption. For the cruise engine, no legitimate Isp figure exists or could exist.
There's a real nuance worth knowing: the drone is launched from its rail via a small solid-fuel rocket-assisted-takeoff (RATO) booster that fires for roughly 2–3 seconds before being jettisoned — that booster technically is a rocket motor with a theoretical Isp, but no manufacturer or OSINT source (Jane's, CSIS, wreckage analyses) has ever published a value for it. It's a disposable, undocumented component.
Verified specs: length 3.5 m, wingspan 2.5 m, mass ~200 kg, cruise speed ~185 km/h, range estimated 1,000–2,500 km (DIA), warhead 40–50 kg standard (a heavier ~90 kg variant reported in later Russian production), GNSS/pre-programmed waypoint guidance. First confirmed Ukraine strike: Sept 13, 2022. A jet-powered derivative, Shahed-238/Geran-3, uses a turbojet and reaches ~600 km/h — turbojets can technically have a specific impulse figure, but none has ever been published for this drone either.
Source: Wikipedia (HESA Shahed 136, Limbach L550E); CSIS Missile Defense Project; militarnyi.com (RATO launch method); Forbes, Sept 2025 (Shahed-238).
| System | Maker / Origin | Guidance | Warhead | Range | Notes |
|---|---|---|---|---|---|
| Shahed-136 / Geran-2 | Iran / Russia (licensed) | GNSS + waypoints | 40–50 kg (up to ~90 kg reported) | 1,000–2,500 km (DIA est.) | Piston engine, ~185 km/h cruise |
| Switchblade 300 (Block 20) | AeroVironment, USA | Man-in-loop, EO/IR | Modular frag / EFP | 30 km | 20+ min endurance, 3.27 lb all-up |
| Switchblade 600 | AeroVironment, USA | Man-in-loop, Javelin-derived seeker | Tandem anti-armor | 55+ mi (90+ km) w/ relay | US Army LASSO program; $990M IDIQ (2024) |
| Harop | IAI, Israel | Man-in-loop, EO/IR, abortable | 15–23 kg (sources vary) | ~200 km effective (datalink-limited) | 6–9 hr endurance; any target type |
| Harpy | IAI, Israel | Fully autonomous, anti-radiation | 32 kg HE | 200 km | No man-in-loop; SEAD/DEAD only |
| Kargu-2 | STM, Turkey | AI-assisted nav/ID, operator-authorized strike | <1.3 kg | ~10 km control link | Center of the disputed 2020 Libya "autonomous kill" claim (see below) |
| RAM-2X | RAM UAV, Ukraine | OSINT-reported, unconfirmed | ~3 kg EFP | ~100 km demonstrated | No official manufacturer datasheet |
| ASN-301 | China (Harpy-derived) | Autonomous, anti-radiation | ~7,000 preformed fragments, laser fuse | 280–300 km | 135–200 kg class |
| CH-901 | China (CASC Rainbow series) | Tube-launched | Small | ~15 km | ~9 kg, up to 2 hr loiter |
The disputed "first autonomous AI kill" claim (Kargu-2, Libya 2020)
A March 2021 UN Panel of Experts report on Libya stated that Kargu-2-class systems were used to "hunt down and remotely engage" retreating forces, which media widely reported as the first AI-autonomous battlefield killing. This is contested: the UN report itself did not confirm the drone was operating autonomously (versus remotely piloted) during any specific engagement, nor definitively confirm a human death resulted. STM's own CEO has stated Kargu-2's autonomy is used mainly for navigation and target differentiation, with an operator required to verify and authorize any strike. Treat the "first AI kill" framing as a widely-repeated but unverified claim, not a confirmed fact.
Source: NPR, Jun 1, 2021; Lieber Institute, West Point, legal/ethical analysis.
Mission Autonomy & Battle-Management Software
The orchestration layer that fuses sensor data and coordinates heterogeneous unmanned systems under one operator. "Lattice OS" is one product in this category, not the category itself.
Lattice (Lattice OS)
Hardware-agnostic platform enabling teams of heterogeneous unmanned systems to collaborate under one human operator. Architecture: ingestion → sensemaking (detection/tracking/correlation) → orchestration (task planning, order routing). Independently corroborated as performing real-time fusion across 100+ sensor types.
Source: anduril.com/lattice/mission-autonomy; Aviation Week; Breaking Defense, Dec 2024.
AIP for Defense / MetaConstellation
AIP for Defense combines LLMs with Palantir's platform for classified-to-tactical-edge AI decision support. MetaConstellation is a distinct product: a satellite-tasking C2 layer requesting imagery/collection across multiple operators' constellations.
Source: palantir.com/platforms/aip/defense/; palantir.com/offerings/metaconstellation/.
Hivemind
AI-pilot autonomy software for GPS/comms-denied environments. Deployed on the V-BAT ISR drone (selected for India's Army) and slated for the X-BAT VTOL fighter; won a USAF CCA production contract for Hivemind, announced June 17, 2026.
Source: shield.ai press releases, 2026.
9Land BMS / Command & Control
Tactical battle management system with a decentralized architecture — loss of individual nodes does not mean loss of functionality — fusing sensor and unit data into a shared digital picture.
Source: saab.com/products/land/command-and-control.
hC2 Software Suite / AMORPHOUS
hC2 is a multi-echelon C2 software suite. AMORPHOUS, unveiled Feb 2025, is swarm-management C2 software designed to command thousands of unmanned systems.
Source: l3harris.com/all-capabilities/hc2-software-suite; Breaking Defense, Feb 2025.
AlphaMosaic
AI decision-support framework born from DARPA's Air Combat Evolution program; cuts weapon-target pairing time from 20–30 minutes to seconds (demonstrated at Eglin AFB, early 2025). Leidos separately holds a $303M USAF ABMS digital-infrastructure contract.
Source: leidos.com/insights/alphamosaic; Air & Space Forces Magazine.
Foreign Air-Defense & Counter-Drone Systems
Non-US systems referenced alongside the above, including the source of a garbled but very real search query.
"Kongsberg anti healing system" — what this search actually means
This is almost certainly a phonetic mis-transcription of "Kongsberg anti-heeling system" — a real Kongsberg Maritime product with nothing to do with air defense or weapons. An anti-heeling system is marine stabilization equipment: ballast tanks are interconnected via automated valves and pumps, and when a heel-angle sensor detects a vessel listing to one side, the system auto-transfers ballast water to the opposite tank to right it, letting cargo operations continue without stopping for manual correction. "Heeling" and "healing" are near-homophones, which is the likely root of the mis-transcription (a marine-parts reseller listing even misspells it the same way).
If the search intent was actually about defense, Kongsberg's real flagship air-defense product is NASAMS (below) — but there's no linguistic path from "anti healing" to NASAMS, so the marine product is the better literal match.
Source: MarineInsight, "What is Anti-Heeling System on Ships"; kongsberg.com fin-stabiliser product line.
NASAMS
Network-centric medium-range ground-based air defense. Each unit: 3 launchers × 6 AMRAAM missiles, can engage up to 72 targets simultaneously. Has continuously defended Washington D.C. airspace since 2005.
- Range ~25 km / ~14–15 km altitude (AIM-120 AMRAAM, NASAMS II)
- Ukraine ~13+ systems (US, Canada, Norway, Lithuania); reported downing of 11 cruise missiles in under 2 minutes, Apr 27, 2025
Source: kongsberg.com; United24 Media, Apr 2025.
Iron Dome
Short-range rocket/mortar interception system. Radar tracks targets 4–70 km; per-interceptor cost is not officially published.
- Intercept range Out to 70 km (vendor)
- Cost estimates $40K–$150K+ per interceptor press estimate only
- Battery cost ~$70–95M, covers 100–150 km²
Source: rafael.co.il; Norsk Luftvern cost analysis, 2025.
Silent Hunter / LW-30 (laser C-UAS)
Fiber-laser system with four power modes (5/10/20/30 kW); interception radius 200 m–4 km depending on mode. Reported (trade-press only) exports to UAE, Saudi Arabia, and Russia.
Source: Wikipedia (Silent Hunter laser weapon); Unmanned Airspace. Export/deployment and combat-performance claims are open-source/trade-press only, not independently confirmed by any government.
UNCONFIRMED DEPLOYMENT CLAIMSCould not verify: any Russian counter-UAS or loitering-munition-defense system with independently verifiable open-source specs meeting this page's sourcing bar.
Glossary
The acronyms and terms used throughout this page, in one place.
Core acronyms: CUAS, ESM, ECM, DRFM, SDR, FPGA, EO/IR, IRST, RATO, C2/BMS, MDS
- CUAS / C-UAS — Counter-Unmanned Aircraft Systems; the broad category of detect/track/classify/engage layers used against hostile drones.
- ESM — Electronic Support Measures: passive detection and geolocation of hostile RF/radar emissions.
- ECM — Electronic Countermeasures: active jamming or deception of RF/radar/comms.
- DRFM — Digital RF Memory: digitally samples and retransmits an altered copy of a radar/RF signal to create false targets.
- SDR — Software-Defined Radio: a radio whose modulation, filtering, and function are set in software rather than fixed hardware.
- FPGA — Field-Programmable Gate Array: reconfigurable hardware logic used for the real-time, nanosecond-latency signal processing ESM/ECM systems need.
- EO/IR — Electro-Optical/Infrared: a sensor combining a visible-light (EO) camera and a thermal/infrared (IR) camera, often on a stabilized gimbal.
- IRST — Infrared Search and Track: passive detection and tracking of targets by their heat signature, without emitting radar.
- RATO — Rocket-Assisted Takeoff: a small, disposable solid-fuel booster used to launch a drone or munition from a rail or truck, jettisoned after a few seconds.
- C2 / BMS — Command and Control / Battle Management System: software that fuses sensor data and coordinates units/effectors under human command.
- MDS — Mission Design Series: the USAF's official aircraft-type naming system (e.g., YFQ-42A) — "Y" prototype, "F" fighter, "Q" unmanned.
- Specific impulse (Isp) — A rocket-engine efficiency metric (thrust per unit propellant weight-flow, in seconds). Does not apply to piston or turbojet engines in the way it's commonly asked about.
Common Mistakes & Misconceptions
The gaps between vendor marketing, press shorthand, and confirmed government data — the errors this research pass had to specifically correct.
Don't conflate vendor claims with confirmed specs
- Directed-energy power figures (Epirus Leonidas, RTX HELWS test-hour claims) are frequently vendor-stated with no independent audit — this page flags each one explicitly.
- CCA speed/range/endurance numbers for YFQ-42A are simply not public; anything you see cited as a hard number for it is speculation.
- Iron Dome's per-interceptor cost has never been officially published — treat every number in circulation as a press estimate.
Program names and designations drift fast
- YFQ-48A is real, but it belongs to Northrop Grumman's Increment 2 contender, not a third Increment 1 aircraft alongside YFQ-42A/44A.
- MQ-28 Ghost Bat is an Australian RAAF/Boeing program, not part of the USAF CCA competition, despite frequent conflation in casual reporting.
- "Anvil," "Lattice," and "Pulsar" are Anduril product names, not category names — the category terms are kinetic interceptor, mission-autonomy software, and RF sensing/jamming node respectively.
Physics and terminology mismatches
- Specific impulse is a rocket-motor metric. Asking for a piston engine's Isp (Shahed-136) is a category error — the right question is horsepower and fuel consumption.
- "First autonomous AI kill" (Kargu-2, Libya 2020) is a widely repeated but formally unconfirmed claim — the underlying UN report does not establish it.
- A mis-transcribed search term ("anti healing") can point to a completely unrelated product line (marine ballast systems) from the same company that also makes air-defense systems — don't assume a defense context by company name alone.