Uncrewed Systems: A Practical Reference

A nano quadcopter weighs between 30 and 500 grams, fits in a pocket, and carries a small electro-optical and thermal camera. It exists to let one soldier look around a corner, into a building, or over a wall without exposing anyone. Endurance is 15–30 minutes, range is a kilometer or two, and the whole system — aircraft, controller, spares, and case — packs into a pouch. Control is manual with stabilization assistance, though newer models add obstacle avoidance and simple return-to-home autonomy.

It is the only aerial system that a dismounted soldier carries and uses without asking anyone's permission, and that immediacy is worth more than its specifications suggest. It flies indoors, hovers to inspect, and is quiet enough at 20–30 m to be hard to notice. The weaknesses are wind, endurance, and range: a 200 g aircraft is unusable above about 25 knots, 25 minutes goes quickly, and the video link is short-ranged and easily jammed. Payload of 50–150 g limits the camera to a small sensor with modest resolution.

Use nano quadcopters for immediate close-in reconnaissance by small units, building clearance, and inspection tasks where a human would otherwise be exposed. If you need to look further than a kilometer or stay up for more than half an hour, a small fixed-wing aircraft is the right tool. If the environment is jammed, expect the video link to be the first thing lost, and prefer models that can complete a preprogrammed route and return without a link. Treat these as consumable — the loss rate in use is high and that is acceptable at this price.

Gross weight 30–500 g · endurance 15–30 minutes · range 1–2 km · payload 50–150 g · wind limit around 25 knots · unit cost a few thousand dollars to roughly $50k.

Control and video links are the weak point, and a handheld jammer at a few hundred dollars defeats most consumer-derived links at short range. GNSS jamming disrupts position hold and return-to-home on models without visual odometry. Small arms work at close range. Simple detection is a challenge, but acoustic and RF detection systems both handle this class reasonably well because the aircraft is loud relative to its size and its link is easy to see.

Teledyne FLIR Black Hornet (the reference nano system, about 33 g), Skydio X2 in its smaller configurations, Parrot ANAFI USA, and the very large number of modified consumer quadcopters in use worldwide.

Consumer drone manufacturing set the price and capability of this class, and defense-specific versions cost five to twenty times as much for hardened links, thermal sensors, and supply-chain assurance. The Black Hornet at roughly $50k per unit and DJI-derived alternatives at a few thousand illustrate the gap, and much of it is procurement policy rather than capability. The supply chain concentration in China remains the central strategic problem for Western buyers of this class, and it has driven substantial investment in domestic alternatives that still cost more.

4 minBlack Hornet 4 Nano UAV – The Future of Battlefield ReconnaissanceTeledyne FLIR · 500k+ views

14 minWhat Can THIS Black Hornet Drone Do?TheRcSaylors · 100k+ views

2 minA first look at the FLIR Black Hornet nanodroneGearScout · 10k+ views

Unmanned Aircraft Systems: Roles, Missions, and Future Concepts (Congressional Research Service) · Joint Rapid Acquisition Cell (US Department of Defense)

A small multirotor ISR drone weighs 1–25 kg, carries a stabilized electro-optical and thermal gimbal, and flies for 25–60 minutes. This is the class that the commercial drone industry built and that every military has adopted, formally or otherwise. Modern examples add substantial onboard autonomy: obstacle avoidance, subject tracking, and the ability to complete a mission without GNSS by using visual-inertial odometry. Payload capacity of 1–5 kg is enough for a good sensor, a small dropped munition, or a radio relay.

Hovering is the capability that matters. A multirotor can hold a position over a target, inspect from any angle, land in a confined space, and launch from anywhere without infrastructure. Cost is low and training time is short. The weaknesses follow from hovering: rotor efficiency in a hover is poor, so endurance stays under an hour regardless of battery improvements, and wind above 25–30 knots ends the mission. Acoustic and visual signature at low altitude is significant, and the aircraft is easy to detect within a kilometer.

Use small multirotors for short-duration overwatch, artillery observation, inspection, and the wide range of tasks where hovering and immediate launch matter more than endurance. If the mission is persistent surveillance of a wide area, a fixed-wing aircraft gives four to ten times the endurance for the same weight. If the requirement is precision strike, a purpose-built loitering munition costs less per effect than modifying an ISR quadcopter. Assume the aircraft will be lost and buy in quantity rather than buying one expensive one.

Gross weight 1–25 kg · endurance 25–60 minutes · payload 1–5 kg · wind limit 25–30 knots · detectable within about 1 km · unit cost $2k–50k.

RF jamming of control and video links is the standard counter and works well against most models. GNSS jamming degrades navigation unless the aircraft carries visual odometry, which the better ones now do. Small arms and shotguns are effective inside 200 m. Detection by acoustic sensors, RF direction finding, and small radar is well developed for this class, and it is the threat that the entire counter-UAS industry was built around.

Skydio X10, Teledyne FLIR SkyRaider, Parrot ANAFI USA, Autel and DJI Mavic and Matrice series in widespread military and paramilitary use, and Ukrainian and Russian modified commercial quadcopters dropping munitions.

This is the most commoditized category in military aviation, with prices from $2k for a modified consumer aircraft to $50k for a hardened Western model. The cost gap has driven a persistent argument about whether supply-chain assurance is worth a 10× price premium, and the practical answer in operational units has often been no. Western manufacturers have gained ground on the back of NDAA compliance requirements, autonomy in GNSS-denied conditions, and cyber assurance, none of which the cheap alternatives match.

3 minSkydio Delivers a Breakthrough for Enterprise Drones with the Launch of Skydio X10Skydio · 500k+ views

13 minSkydio X10 Hands On First Look - A Huge UpgradeBilly Kyle · 50k+ views

18 minParrot Anafi USA First Flight and Impressions - Small But MightyBilly Kyle · 10k+ views

Unmanned Aircraft Systems: Roles, Missions, and Future Concepts (Congressional Research Service) · Unmanned Aircraft Systems (Federal Aviation Administration)

A hand-launched fixed-wing drone weighs 2–20 kg, is thrown or bungee-launched by one or two people, and flies for one to four hours at 30–60 knots. Recovery is by belly landing, parachute, or a deep stall into a net. The wing does the work that a multirotor's rotors have to do continuously, so endurance is four to ten times better for the same battery. Range from the control station is typically 15–50 km, extended further by autonomous waypoint flight with periodic link contact.

Endurance and area coverage are the advantages, and they are large: a 4 kg fixed-wing aircraft covers in one sortie what a quadcopter would need six sorties to see. It is also quieter at altitude and harder to detect than a multirotor. The weaknesses are launch and recovery, which need space and produce wear, and the inability to hover — you get a pass over the target rather than a stare. Wind tolerance is better than a multirotor's but landing in a confined space is difficult, and airframe damage on recovery is the main maintenance driver.

Pick hand-launched fixed-wing aircraft for battalion and company-level ISR over routes and areas, for artillery observation at range, and for communications relay. If the task requires hovering over a specific point or operating inside a building, use a multirotor. If the requirement is more than four hours or a heavier sensor, a Group 3 catapult-launched system is the next step up in both capability and cost. A hybrid VTOL fixed-wing splits the difference at the cost of complexity and about 20% of the endurance.

Gross weight 2–20 kg · endurance 1–4 hours · cruise 30–60 knots · control range 15–50 km · typical operating altitude 300–500 m · unit cost $20k–100k.

Control links are jammable, though these aircraft usually complete their route autonomously if the link drops. GNSS jamming is more significant, and models without terrain or visual navigation will fail to return. They are small and quiet enough that visual and acoustic detection at altitude is hard, but RF detection of the datalink works. Small arms fire is largely ineffective at typical operating altitudes of 300–500 m.

AeroVironment RQ-11 Raven and RQ-20 Puma, Elbit Skylark, the Ukrainian Leleka-100 and Furia, WB Group FlyEye, and a very large number of similar systems from many manufacturers.

This class costs $20k–100k per aircraft and has a broad international supplier base with low barriers to entry, since the airframe is composite and the electronics are commercial. Attrition in active use is high — measured in weeks of operational life in contested areas — which has shifted buying patterns from small numbers of expensive systems toward larger numbers of cheaper ones. The systems that hold their value do so through autonomy in GNSS-denied conditions rather than through airframe quality.

6 minRQ-11 Raven UAV Drone - Hand Thrown Launch & Break-apart Landing.AiirSource Military · 100k+ views

2 minRQ-11B Raven - Soldiers assembling and launching UAVMilitaryNotes · 10k+ views

2 minRQ-20 Puma Unmanned Air Vehicle (UAV) launching in Forward Operating Base | AiirSourceAiirSource Military · 10k+ views

Unmanned Aircraft Systems: Roles, Missions, and Future Concepts (Congressional Research Service) · Joint Rapid Acquisition Cell (US Department of Defense)

An FPV attack drone is a racing quadcopter carrying an explosive charge, flown into its target by a pilot watching a video feed through goggles. Airframes are 7 to 15 inches across, cost $300–1,500 including the warhead, and carry 0.5–3 kg of explosive out to 5–20 km. The whole concept depends on cheap components — brushless motors, lithium packs, analog or digital video links — from the consumer hobby market, and on a trained pilot rather than a guidance system. It has become the most consequential weapon of the current war in Ukraine by volume of targets destroyed.

Cost per effect is the entire argument. A $500 drone destroying a $3M tank is a cost exchange that no traditional weapon matches, and production can run to hundreds of thousands of units per year from workshop-scale facilities. Precision is excellent because a human flies it to a specific point on the target. The weaknesses are range, weather, and the pilot. Twenty kilometers is the practical limit, wind and rain ground them, each aircraft needs a trained operator for its entire flight, and the video and control links are exactly what electronic warfare is designed to attack.

Use FPV drones for short-range precision attack against vehicles, positions, and infantry where a human can identify the aimpoint and the target is inside 20 km. If the target is further away, a fixed-wing one-way attack drone is the right tool. If the electromagnetic environment is heavily contested, fiber-optic control or terminal autonomy is necessary and radio-controlled FPVs will mostly fail. The doctrinal implication is larger than the tactical one: this class has made the last 10 km of ground a very dangerous place for any vehicle, and it has changed how armies disperse and move.

Airframe 7–15 inches across · warhead 0.5–3 kg · range 5–20 km · unit cost $300–1,500 including the warhead · production hundreds of thousands to millions of units per year.

RF jamming of the control and video links is the primary counter and has been effective enough to drive both sides toward fiber-optic control and terminal autonomy. Cage armor and slat screens defeat many shaped-charge warheads. Electronic warfare at the vehicle level, drone-on-drone interception, and shotguns all contribute. The countermeasure race here is the fastest-moving in modern warfare: each counter has held for a few months before a workaround appeared.

Ukrainian and Russian workshop-produced FPV drones in enormous quantities, commercial platforms from Chinese suppliers adapted for the role, Western entrants such as Neros and Performance Drone Works, and the fiber-optic variants that emerged in 2024 to defeat jamming.

This is manufacturing at a scale and price point that traditional defense industry has never operated at: unit costs of a few hundred dollars, production measured in millions of units per year, and design iteration in weeks. Ukraine's stated production targets have run into the millions annually. The lesson that Western procurement is currently absorbing is that the relevant metric is drones per month per dollar, and that a system taking three years to qualify is obsolete before it fields.

14 minDrone Theory 101: Part 1. The basics, and how an fpv quadcopter functions!Riley Morgan · 500k+ views

10 minFPV Drone Component Breakdown [FC, ESC, Motors, Camera, VTX, Frame...]Rotor Riot · 50k+ views

10 minUkrainian Kamikaze FPV Drones ExplainedNarekFPV · 1k+ views

Emerging Military Technologies: Background and Issues for Congress (Congressional Research Service) · Tactical Developments During the Third Year of the Russo–Ukrainian War (Royal United Services Institute)

A VTOL fixed-wing hybrid carries both lift rotors for takeoff and landing and a wing plus a pusher propeller for cruise. It takes off vertically from any small clearing, transitions to wing-borne flight, and lands vertically at the end. Typical systems weigh 5–25 kg, fly for two to six hours, and carry a stabilized sensor. The lift rotors are dead weight and drag in cruise, which costs roughly 15–25% of the endurance a pure fixed-wing aircraft of the same weight would achieve, and that is the price of eliminating launch and recovery equipment.

Operational simplicity is the advantage. No catapult, no net, no runway, no recovery damage, and one or two people can operate the system from a clearing. That removes the logistics and the airframe attrition that hand-launched and catapult-launched systems suffer on recovery. The weaknesses are complexity and endurance. There are more motors and more failure modes, the transition between hover and wing-borne flight is the most demanding part of the flight envelope, and endurance sits below a comparable pure fixed-wing aircraft. Hybrid combustion-electric versions push endurance to 8–12 hours at higher cost.

Pick VTOL fixed-wing when the operating site has no room or time for launch and recovery equipment and you still need multi-hour endurance: maritime operations from small vessels, mountainous or forested terrain, and units that move frequently. If you have space for a catapult and a net, a conventional fixed-wing aircraft gives more endurance per dollar. If the mission is under an hour, a multirotor is simpler and cheaper. The recovery problem is usually what decides this, and it is worth checking how the system behaves when it has to land in wind on a moving deck.

Gross weight 5–25 kg · endurance 2–6 hours electric and 8–12 hours hybrid · endurance penalty 15–25% against a pure fixed-wing of the same weight · unit cost roughly 1.5–2× a comparable fixed-wing system.

The same counters as other small UAS apply: link jamming, GNSS denial, and detection by RF and acoustic means. The transition phase is the most vulnerable point mechanically, and systems that lose a lift motor during transition are usually lost. Because these aircraft are more expensive than hand-launched ones, attrition matters more, and units tend to fly them more conservatively than the mission would ideally allow.

Quantum Systems Vector and Trinity, Threod Systems and Delair platforms, WingtraOne in survey use, Silvertone and Insitu VTOL variants, and a growing number of Ukrainian and Chinese designs in the same class.

VTOL fixed-wing designs cost roughly 1.5–2× a comparable pure fixed-wing system, and the market has grown quickly because the operational savings from eliminating launch and recovery equipment usually exceed that premium. Component costs are commercial, so barriers to entry are low and the field is crowded. Differentiation comes from transition reliability, wind tolerance, and GNSS-denied navigation rather than from airframe design, which has largely converged.

11 minHow does a VTOL UAV work?AiTelly · 50k+ views

14 minHow do Autonomous VTOL Drones Work? QuadPlanes ExplainedPathfinder George · 10k+ views

5 minALTi Transition VTOL fixed-wing drone flies for 6 hoursDrone & Sundry · 100k+ views

Unmanned Aircraft Systems: Roles, Missions, and Future Concepts (Congressional Research Service) · Greased Lightning (GL-10) Flight Testing Campaign (NASA Technical Reports Server)

A fiber-optic guided drone trails a spool of single-mode optical fiber from the aircraft or the launch point, carrying video down and commands up through glass rather than through the air. Spools hold 10–25 km of fiber weighing a few hundred grams. The aircraft is otherwise an ordinary FPV quadcopter. The point is simple: a fiber link cannot be jammed, cannot be direction-found, and works perfectly inside buildings, in tunnels, and behind terrain where radio does not reach. It appeared at scale in Ukraine in 2024 as a direct answer to electronic warfare.

Immunity to jamming is decisive in an environment where radio-controlled drones fail most of the time. The link also gives no RF emission, so the operator's position is not revealed by direction finding, which is a significant survivability gain. Video quality is better than a compressed radio link. The weaknesses are physical: the fiber snags on trees and structures, it can break, the spool limits range to about 25 km, the aircraft carries several hundred grams of dead weight, and flight profiles must avoid sharp maneuvers that would kink the fiber. Spent fiber also litters the ground and marks routes.

Use fiber-optic control where electronic warfare has made radio links unreliable, where the operator must not be located by direction finding, and where the route allows a fiber to pay out without snagging — open terrain, roads, and urban approaches from a stable direction. If the target is beyond 25 km, or the route is heavily wooded, use a radio-controlled or autonomous system instead. The broader lesson is that a physical link is currently the most effective answer to jamming, and defenders have had to shift from electronic counters to physical interception.

Fiber spool 10–25 km · spool weight a few hundred grams · practical range limit about 25 km · spool cost $150–500 · airframe a few hundred dollars before the spool.

The fiber can be cut, and deliberate wire-cutting measures have appeared. Snagging on vegetation and structures ends missions without any enemy action. The spool limits range absolutely. Because there is no RF to detect, defenders have moved to acoustic and optical detection plus interceptor drones and small-arms fire, which is a harder and more expensive problem than jamming was. Spent fiber on the ground also reveals attack routes and operator positions to observant defenders.

Russian Knyaz Vandal Novgorodsky and Prince Vandal series, Ukrainian fiber-optic FPV production from multiple workshops, and Chinese-supplied spools that constitute most of the world's supply of suitable fiber and dispensers.

A fiber spool adds $150–500 to a drone that otherwise costs a few hundred dollars, so the total remains far below any guided munition. Supply is the interesting constraint: the specific fiber and dispenser designs come overwhelmingly from a small number of Chinese producers, which makes this a concentrated dependency for both sides in the current conflict. Expect Western programs to treat spool production as a supply-chain priority if this approach persists.

7 minFiber Optic FPV Drones? What Does Bardwell Think? - FPV QuestionsJoshua Bardwell Livestream Clips · 10k+ views

14 minWhat are Fiber Optic FPV Drones, and Why are They so Deadly? - Feat. @DefenseTechforUkraineNorth Wind Aerial · 5k+ views

2 minHow to connect the FPV drone with fiber optic canister to the air and ground unitJackie zhou · 5k+ views

The Russia-Ukraine Drone War: Innovation on the Frontlines and Beyond (Center for Strategic and International Studies) · Emerging Military Technologies: Background and Issues for Congress (Congressional Research Service)

A catapult-launched tactical UAS weighs 20–80 kg, is fired off a pneumatic or hydraulic rail, and is recovered by flying into a vertical wire (the SkyHook method) or a net. Endurance runs 12–24 hours with a small piston or heavy-fuel engine, and the payload is a stabilized electro-optical and infrared turret plus often a communications relay or signals-intelligence package. This class is the workhorse of persistent tactical surveillance, particularly at sea, because the launch and recovery equipment fits on a ship's deck without a flight deck.

Endurance per dollar is the strength. Twenty-plus hours on station from a 20 kg aircraft costing a few hundred thousand dollars is a very efficient way to watch a piece of ground or ocean. Heavy-fuel engines allow shipboard operation on the same fuel as everything else. The weaknesses are the launch and recovery equipment, which is bulky, needs setup and maintenance, and is a significant fraction of the system cost, and the small payload, which caps sensor quality. Recovery in high sea states or strong crosswinds is genuinely difficult and is where most losses occur.

Pick this class for persistent maritime and land surveillance where 12–24 hours on station matters and there is no runway: ship-based ISR, border and maritime patrol, and brigade-level overwatch. If you have a runway, a conventional Group 3 aircraft carries more payload for the same money. If you need vertical launch and recovery from a very small deck, a rotary or VTOL system is easier to operate at the cost of endurance. Budget for the launcher and recovery system as part of the buy — they often cost as much as the aircraft.

Gross weight 20–80 kg · endurance 12–24 hours · operating altitude 5,000–15,000 ft · aircraft cost $200k–1M · full system with launcher and recovery gear $5–20M.

These aircraft operate at 5,000–15,000 ft and are within reach of most short-range air defense and man-portable systems. Their datalinks are detectable and jammable, though they complete routes autonomously. Because they loiter predictably over an area for many hours, they are relatively easy to detect and to plan against. Losses in contested airspace have been high in recent conflicts, which is the main reason this class is increasingly treated as attritable rather than as an asset to protect.

Insitu ScanEagle and Integrator (RQ-21 Blackjack), Elbit Skylark 3, Aeronautics Orbiter, Textron Aerosonde, and the Chinese and Turkish equivalents that have proliferated widely.

Aircraft cost $200k–1M and full systems with launchers, recovery equipment, and ground stations run $5–20M. The business model has historically included contractor-operated services sold by the flight hour, which is a substantial revenue stream. Competitive pressure now comes from cheaper fixed-wing systems from Ukraine, Turkey, and China that deliver 60–70% of the capability at a fifth of the price, and from VTOL designs that eliminate the launch and recovery equipment entirely.

3 minInsitu ScanEagle Launch And CaptureAIRBOYD · 100k+ views

2 minLanding Drone Aircraft With No Runway: Unique 'SkyHook' Recovery SystemAiirSource Military · 500k+ views

4 minScanEagle UAV Launch & RecoveryAiirSource Military · 10k+ views

Unmanned Aircraft Systems: Roles, Missions, and Future Concepts (Congressional Research Service) · Joint Rapid Acquisition Cell (US Department of Defense)

A runway-launched tactical UAS in the Group 3 class weighs 150–600 kg, takes off and lands conventionally on a short strip, and flies for 12–20 hours at 10,000–20,000 ft. Payload of 25–150 kg allows a good sensor turret plus a synthetic aperture radar or a signals-intelligence package, and armed variants carry small precision munitions. This is the class where Turkey's Bayraktar TB2 changed the market, demonstrating that a $5M aircraft with a $70k missile could destroy air-defense systems and armored vehicles in a way that had previously required far more expensive assets.

Capability per dollar is the strength: a modest aircraft carrying a good sensor and two to four small munitions delivers most of the useful output of a much larger armed drone. Operating cost is a fraction of a crewed aircraft's. The weaknesses are survivability and infrastructure. These aircraft are slow, non-stealthy, and fly predictably, so a functioning integrated air defense system removes them quickly — TB2 losses rose sharply in Ukraine once Russian air defenses were organized. They also need a prepared strip of 500–1,000 m, which is an infrastructure and vulnerability consideration.

Pick this class for armed ISR against opponents without effective medium-range air defense, for border and maritime patrol, and for persistent surveillance where a runway is available. Against a peer with layered air defense, plan on losing them and price accordingly, or use them outside the threat envelope as sensor and communications platforms. The strategic lesson from 2020–2022 is that this class is decisive against unprepared opponents and attritable against prepared ones, and both conclusions are correct.

Gross weight 150–600 kg · endurance 12–20 hours · operating altitude 10,000–20,000 ft · payload 25–150 kg · runway 500–1,000 m · unit cost roughly $5M for a TB2 against $30M+ for a Western equivalent.

Medium-range surface-to-air missiles and modern short-range systems engage them readily. Electronic warfare against control links and GNSS is effective, particularly since most of these aircraft rely on satellite or line-of-sight links for weapons employment. They are large enough and slow enough to be tracked by ordinary surveillance radar. Their runway requirement also makes their operating bases predictable and targetable.

Baykar Bayraktar TB2 and Akinci, General Atomics Gray Eagle in its smaller configurations, Textron Shadow, IAI Heron, CAIG Wing Loong I, and a large number of similar designs from Turkish, Chinese, and Iranian manufacturers.

The TB2 at roughly $5M per aircraft against $30M+ for a comparable Western system reset expectations for the whole category, and Turkish and Chinese suppliers now dominate export sales on price. Operating cost is a few thousand dollars per flight hour against tens of thousands for crewed aircraft. Western manufacturers have responded partly on capability and partly on regulatory arguments, but the price gap has been the decisive factor in most competitions outside NATO.

6 minHow a Military Drone Works | Bayraktar TB2 UAVAiTelly · 1m+ views

10 minThe Truth About Ukraine's Bayraktar TB2 Drone: Project UkraineTask & Purpose · 500k+ views

4 minTurkish Drone Technology | Making of BAYRAKTAR TB2 UAVTR · 10k+ views

Unmanned Aircraft Systems: Roles, Missions, and Future Concepts (Congressional Research Service) · Publications (Stockholm International Peace Research Institute)

A tactical VTOL UAS weighs 100–600 kg and takes off and lands vertically, either as a multirotor, a tilt-rotor, or a fixed-wing aircraft with dedicated lift fans. Endurance runs 6–16 hours with a heavy-fuel piston or turbine engine, and payload capacity of 25–150 kg supports a proper sensor turret or a useful cargo load. The category has grown quickly because it removes the launch and recovery equipment that constrains where a Group 3 system can operate, and because ships and forward units both want aviation without a flight deck or a runway.

Operating from anywhere is the point: a small ship's deck, a clearing, a road. That eliminates catapults, nets, and runways along with their cost, footprint, and vulnerability. Cargo variants can resupply positions that a helicopter would be too valuable or too loud to reach. The weaknesses are efficiency and complexity. Vertical lift costs endurance relative to a conventional aircraft of the same weight, hybrid configurations carry dead weight in cruise, and the transition envelope adds failure modes. Downwash and acoustic signature during landing are significant.

Pick tactical VTOL where the operating site cannot support conventional launch and recovery and the mission needs more than a small drone can deliver: small-deck maritime operations, expeditionary units, and forward resupply. If a runway or a catapult is available, a conventional Group 3 aircraft gives more endurance and payload per dollar. For pure cargo missions, compare against ground resupply seriously — an uncrewed helicopter delivering 200 kg is impressive but often more expensive per tonne than a truck when the route is passable.

Gross weight 100–600 kg · endurance 6–16 hours · payload 25–150 kg · cargo variants around 200 kg · unit cost $1–15M, rotary designs at the high end.

The same air defense and electronic warfare counters as other Group 3 systems apply. The hover phases during takeoff and landing are acoustically and visually conspicuous, and they are also the least recoverable part of the flight if a lift unit fails. Cargo variants flying predictable routes to known positions are straightforward to ambush, which is a tactical planning issue rather than a technical vulnerability.

Northrop Grumman MQ-8C Fire Scout, Schiebel Camcopter S-100, UMS Skeldar V-200, Airbus VSR700, Kaman KARGO and the K-MAX unmanned cargo helicopter, and a growing number of Chinese and Korean designs.

Aircraft costs run $1–15M depending on size and payload, with rotary designs at the high end. Naval demand drives much of the market, since navies want organic aviation on ships too small for a helicopter. The uncrewed cargo segment has struggled commercially because the economics rarely beat ground transport except in genuinely contested or inaccessible conditions, which is a smaller market than early forecasts assumed.

4 minJUMP 20 VTOL Fixed-Wing Medium Unmanned Aircraft System | Explainer VideoAV · 100k+ views

8 minLatest on Shield AI's MQ-35 V-BAT drone and Hive Mind AI autonomy software.TWZ · 10k+ views

4 minKaman K-MAX Helicopter • Cargo Resupply Unmanned AircraftGung Ho Vids · 10k+ views

Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service) · Modular Unmanned Aerial System with Multi-Mode Propulsion (NASA Technical Reports Server)

A ship-launched rotary UAS is a purpose-built or converted unmanned helicopter operating from a warship's flight deck, typically weighing 200–3,000 kg. It carries a maritime surveillance radar, an electro-optical turret, and often a sonobuoy dispenser or a signals-intelligence package, and it flies for 8–14 hours. Deck handling — securing the aircraft in a rolling sea, refuelling, and moving it into a hangar — is a substantial part of the engineering, and it is where uncrewed operation is hardest, because there is no pilot to make judgment calls during recovery.

It extends a small warship's sensor horizon from about 20 nautical miles to 100 or more without risking a crew, and it does so for a fraction of a crewed helicopter's operating cost. Persistence is far better than a crewed aircraft, since there is no crew duty cycle. The weaknesses are deck operations and cost. Automated deck landing in sea state 5 is demanding, hangar space on small ships is contested, and the aircraft themselves are expensive enough that they are not treated as attritable. Maintenance load per flight hour is closer to a helicopter's than to a small drone's.

Pick a ship-launched rotary UAS when a warship needs persistent organic surveillance beyond its radar horizon and has deck space but not enough for a full helicopter detachment. If the ship can host a crewed helicopter and the mission needs lifting or boarding capability, the crewed aircraft is more versatile. If the requirement is simply extending the sensor horizon at low cost, a small catapult-launched fixed-wing system does it for a tenth of the price with less deck impact.

Gross weight 200–3,000 kg · endurance 8–14 hours · sensor horizon 100+ nautical miles against about 20 from the ship's own radar · automated deck landing demanding above sea state 5 · unit cost roughly $9–12M per aircraft at MQ-8C size, well below that for the 200 kg class.

The aircraft is a conventional radar target and is engageable by any naval or shore-based air defense. Its datalink to the ship is jammable, though modern systems are designed to continue autonomously and return. The operational limit is more often sea state and deck availability than enemy action. Because it operates from a known ship, its presence is predictable.

Northrop Grumman MQ-8B and MQ-8C Fire Scout, Schiebel Camcopter S-100 in service with several navies, Saab and UMS Skeldar systems, Leonardo AWHero, and Chinese naval rotary UAS on newer frigates.

Programs in this category have had a mixed record: the MQ-8B and C were procured and then retired early by the US Navy, which concluded that the capability did not justify the deck footprint and sustainment cost on littoral combat ships. Navy production contracts for the MQ-8C worked out to roughly $9–12M per aircraft. Smaller systems like the S-100 have done better commercially by fitting more ships and costing less. The general lesson is that deck space on a warship is extraordinarily expensive real estate, and a system has to justify it against every alternative use.

16 minThe MQ-8 Fire Scout: America's Unmanned Robot HelicopterMegaprojects · 100k+ views

2 minNorthrop Grumman-built MQ-8C Fire Scout First FlightNorthrop Grumman · 10k+ views

3 minSchiebel CAMCOPTER® S-100 UAS - Maritime (Canada - Fogo Island)Schiebel · 5k+ views

Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service)

A medium-altitude long-endurance armed UAS weighs 1,500–5,000 kg, cruises at 20,000–30,000 ft, and stays airborne for 24–40 hours. It carries a large sensor turret, a synthetic aperture radar, signals-intelligence payloads, and 500–1,700 kg of weapons. Control is via satellite link from a ground station that may be thousands of kilometers away, with a launch and recovery element near the operating base. This is the class that defined two decades of counterinsurgency air operations, and its central capability is not weapons but persistence — the ability to watch one place continuously for a day and a half.

Persistence plus precision strike from one platform is the strength, and it is genuinely transformative against targets that must be observed before they can be attacked. Operating cost is roughly $3,000–5,000 per flight hour against $20,000+ for a crewed strike aircraft. The weaknesses are survivability and bandwidth. These aircraft are large, slow, and non-stealthy, so they cannot operate where an opponent has functioning air defenses. Each aircraft consumes substantial satellite bandwidth, and that has been a real constraint on how many can operate at once.

Use MALE UAS where you have air superiority and need persistent observation and prompt strike: counterinsurgency, counterterrorism, border surveillance, and maritime patrol. Against a peer opponent with layered air defense, they cannot survive over defended territory and are better used as standoff sensor and communications platforms outside the threat ring. If the requirement is purely surveillance at high altitude over a wide area, a HALE aircraft covers far more ground; if it is a specific short-notice target, a loitering munition may be a cheaper way to get the same effect.

Gross weight 1,500–5,000 kg · cruise altitude 20,000–30,000 ft · endurance 24–40 hours · weapons load 500–1,700 kg · operating cost $3,000–5,000 per flight hour · unit cost roughly $30M for an MQ-9.

Any modern medium-range surface-to-air missile removes them, and several have been shot down by short-range systems and even by fighters. Satellite control links are jammable, and disrupting them either ends the mission or forces autonomous return. GNSS denial degrades weapons employment. Their basing is well known and their orbits are predictable, so planning against them is not difficult for a capable opponent.

General Atomics MQ-9 Reaper and MQ-1C Gray Eagle, Turkish Aerospace Anka, CAIG Wing Loong II, Chengdu CH-4 and CH-5, IAI Heron TP, and the European Eurodrone program.

An MQ-9 costs roughly $30M per aircraft with a system price several times that including ground stations, and export is governed by the Missile Technology Control Regime, which historically restricted US sales and opened the market to Chinese and Turkish suppliers who face no such constraint. That policy choice reshaped the global armed-drone market more than any technical factor. The category now faces pressure from both directions: cheaper Group 3 systems from below, and collaborative combat aircraft designed for contested airspace from above.

14 minMQ 9 Reaper Drone ExplainedAiTelly · 500k+ views

7 minMQ-9 Reaper and Its Weapon Systems: A Detailed Explainer 3D AnimationEXPLANATION-AVENUE - 3D Animations by Oliver Ende · 100k+ views

17 minHow Air Force Drone Pilots "Fly" The $32 Million MQ-9 Reaper Drone | Boot Camp | Business InsiderBusiness Insider · 1m+ views

U.S. Unmanned Aerial Systems (Congressional Research Service) · Publications (Stockholm International Peace Research Institute)

A high-altitude long-endurance UAS operates at 55,000–65,000 ft for 24–34 hours, above weather and above most air traffic. From that altitude the sensor horizon extends roughly 500 km, so one aircraft surveys enormous areas — a single Global Hawk sortie can image tens of thousands of square kilometers with synthetic aperture radar. The airframe is a very high aspect ratio wing on a light structure, optimized entirely for altitude and endurance, which makes it fragile on the ground and sensitive to weather during climb and descent through the lower atmosphere.

Area coverage per sortie is unmatched by anything except satellites, and unlike a satellite the aircraft can be retasked in flight and can stare at one place for a day. Operating from above the weather makes optical sensing reliable. The weaknesses are cost, fragility, and vulnerability. Unit costs run over $100M, the aircraft is slow and highly visible on radar, and it is defenseless — a high-altitude surface-to-air missile engages it easily, as the 2019 loss of an RQ-4 to an Iranian SAM demonstrated. Weather during the long climb and descent is a genuine operational constraint.

Use HALE aircraft for wide-area surveillance in permissive or semi-permissive airspace: maritime domain awareness, treaty monitoring, disaster response, and standoff collection along a border. In contested airspace they cannot survive, and the alternative is satellite collection or standoff sensing from outside the threat envelope. Compare seriously against commercial satellite imagery and RF geolocation services, which now provide a substantial fraction of the wide-area picture at a fraction of the cost, though without the ability to stare.

Operating altitude 55,000–65,000 ft · endurance 24–34 hours · sensor horizon roughly 500 km · coverage tens of thousands of square kilometers per sortie · unit cost over $100M.

Long-range surface-to-air missiles engage them without difficulty. They are large, slow radar targets with no self-protection. Satellite links are jammable. Weather at the operating base constrains launch and recovery more than most aircraft. Their high cost and small numbers mean that losing even one is strategically significant, which constrains how aggressively they are employed.

Northrop Grumman RQ-4 Global Hawk and MQ-4C Triton, the Euro Hawk and NATO AGS variants, Chinese WZ-7 Soaring Dragon, and high-altitude pseudo-satellite programs such as Airbus Zephyr that pursue the same coverage with solar-electric aircraft at much lower cost per hour.

Program costs have been very high and the value proposition has been repeatedly challenged. The US Air Force sought to retire Global Hawk blocks partly because commercial satellite imagery covered much of the same requirement more cheaply. Solar-electric pseudo-satellites promise weeks of endurance at far lower operating cost, and while they carry much smaller payloads, they are the most credible economic challenger to this category.

9 minHow Northrop Grumman's High-Altitude Unmanned Aircraft WorkNorthrop Grumman · 1k+ views

20 minRQ-4 Global Hawk: The Future of WarfareSkyships Eng · 10k+ views

2 minUnmatched Surveillance with MQ-4C TritonNorthrop Grumman · 50k+ views

U.S. Unmanned Aerial Systems (Congressional Research Service) · Supporting NASA Science with High-Altitude Long-Endurance Aircraft (NASA Technical Reports Server)

A stealth uncrewed combat air vehicle is a low-observable flying-wing aircraft designed to penetrate defended airspace, carrying weapons and sensors internally. Removing the cockpit removes the canopy — a significant radar and infrared feature — and removes the g-limits and life-support volume that shape crewed aircraft, so the airframe can be optimized purely for signature and range. Demonstrators have proven the hard parts, including autonomous carrier launch and recovery and autonomous aerial refuelling, which the X-47B did in 2013 and 2015 respectively.

Survivability in contested airspace is the point, and it is the capability that MALE drones lack entirely. Endurance far exceeds a crewed aircraft's because there is no crew to fatigue, and unrefuelled combat radius of 1,500–2,000 km is achievable. The weaknesses are cost and command. These aircraft cost as much as crewed fighters, so they are not attritable in any meaningful sense, and they need a control approach that works when satellite links are jammed — which in a contested environment means substantial autonomy, and that raises both technical and policy questions that remain unresolved.

A stealth UCAV is the right answer when the mission requires penetrating a modern integrated air defense system to collect or to strike, and putting a crew at that risk is unacceptable or unnecessary. If the airspace is permissive, a MALE aircraft does the job at a tenth of the cost. If the target can be attacked from standoff range, a cruise missile is cheaper per shot and does not need to come home. The category's real competition is a mix of standoff weapons and collaborative combat aircraft, and program history suggests that competition has usually won.

Gross weight roughly 7,000–20,000 kg across demonstrators · unrefuelled combat radius 1,500–2,000 km · internal payload around 2,000 kg on the larger types · unit cost comparable to a crewed fighter · autonomous carrier launch and recovery demonstrated in 2013, autonomous aerial refuelling in 2015.

Low-frequency and passive radar detect low-observable aircraft at useful ranges, though usually without engagement-quality accuracy. Infrared search and track systems detect them passively. Satellite control links are jammable, which forces reliance on onboard autonomy. Above all, the counter is economic: a small number of very expensive aircraft can be attrited faster than they can be replaced, which is the argument that has repeatedly pushed programs toward cheaper collaborative aircraft instead.

Northrop Grumman X-47B (the carrier demonstrations), Boeing X-45, Dassault nEUROn, BAE Taranis, the Russian S-70 Okhotnik, and the Chinese GJ-11 Sharp Sword. The US Navy's UCLASS program was restructured into the MQ-25 tanker, which is the clearest signal of how the cost argument resolved.

Every major stealth UCAV program has been canceled, restructured, or slowed. The pattern is consistent: the aircraft ends up costing as much as a crewed fighter, which removes the cost argument for uncrewed operation, while the autonomy and command-and-control questions remain hard. That history is why the current generation of programs has pivoted to cheaper collaborative aircraft designed to be lost, rather than exquisite penetrating platforms designed to survive.

4 minX-47B UCAS Aviation History Under WayNorthrop Grumman · 1m+ views

2 minX-47B Completes First Autonomous Aerial RefuelingNorthrop Grumman · 50k+ views

3 minDassault nEUROn European UCAV Drone Prototype Takeoff And Landing #1Siivet - Wings · 50k+ views

U.S. Unmanned Aerial Systems (Congressional Research Service)

A collaborative combat aircraft is a jet-powered uncrewed aircraft designed to fly alongside crewed fighters, carrying weapons, sensors, or jammers, and taking direction from the crewed aircraft rather than from a distant ground station. The design target is a unit cost of a quarter to a third of a crewed fighter — the US Air Force has spoken about $25–30M — achieved by accepting less stealth, less range, and a shorter service life. The concept accepts that some aircraft will be lost on every mission and prices them accordingly.

Mass at acceptable cost is the argument. Adding two to five uncrewed aircraft to each crewed fighter multiplies the sensors, weapons, and jammers in a formation without multiplying the pilots, and it forces an opponent to allocate defensive resources against many more objects. The weaknesses are autonomy and command. In a jammed environment the aircraft must decide a great deal for itself, which is both a hard engineering problem and an unresolved policy one. Sustainment for a large fleet of jet aircraft is also expensive regardless of how cheap the airframes are.

Collaborative aircraft are the current answer to fighting in contested airspace at acceptable cost, and they are most valuable where an opponent's air defense makes crewed penetration expensive. If the airspace is permissive, cheaper systems do the job. If the requirement is one specific target, a standoff missile is cheaper per effect. The pacing questions for any program in this class are the achieved unit cost, the autonomy behavior when the link is lost, and whether basing and sustainment for hundreds of jet aircraft is actually affordable.

Target unit cost $25–30M, roughly a quarter to a third of a crewed fighter · 2–5 uncrewed aircraft per crewed fighter · endurance several hours to about a day · planned US Air Force buys in the hundreds to over a thousand airframes.

Air defense engages them like any other aircraft, and their reduced stealth relative to a crewed fighter makes that easier. Electronic attack on the crewed-to-uncrewed link degrades coordination, which is exactly why autonomy matters. Because the concept depends on numbers, the real counter is economic: if interceptors or air defense missiles cost less than the aircraft they destroy, the exchange favors the defender and the concept weakens.

Anduril YFQ-44A Fury and General Atomics YFQ-42A under the US Air Force CCA program, Boeing MQ-28 Ghost Bat with the Royal Australian Air Force, Kratos Valkyrie XQ-58, the European Remote Carrier concepts within FCAS, and the Turkish Kizilelma.

This is currently the most heavily funded area in uncrewed aviation, and the central bet is that unit cost can be held low enough to buy hundreds. History with the stealth UCAV programs suggests that is the hard part: requirements creep pushes cost up, and once the aircraft costs half a fighter the concept loses its rationale. The programs to watch are the ones enforcing a cost ceiling as a hard requirement rather than treating it as a goal.

19 minThe Air Force's Robot Wingmen Are HereTask & Purpose · 50k+ views

3 minMQ-28, Wedgetail, Super Hornet: Drone Intercept Behind-the-ScenesBoeing · 100k+ views

18 minKratos XQ-58 Valkyrie: A Loyal Wingman for the US Marine CorpsMegaprojects · 100k+ views

U.S. Air Force Collaborative Combat Aircraft (CCA) (Congressional Research Service)

Uncrewed cargo aircraft carry freight without a crew, either as purpose-built designs or as conversions of existing light transports with an autonomy kit installed. Payloads range from a few hundred kilograms to several tonnes. The commercial case is regional air freight where pilot cost and availability are the binding constraints; the military case is resupply into places where a crewed aircraft would be at unacceptable risk, and aerial refuelling, where the MQ-25 is the first operational uncrewed tanker.

Removing the crew removes crew cost, crew rest limits, and crew risk, which is worth a great deal on routes that are dull, long, or dangerous. Conversion kits let operators use existing certified airframes, which shortcuts much of the development. The weaknesses are regulatory and economic. Certifying an uncrewed aircraft to operate in civil airspace remains slow and unfinished in most jurisdictions. And in ordinary conditions the economics are marginal: crew cost is a modest fraction of total operating cost on most cargo routes, so removing it does not transform the business the way early projections suggested.

Use uncrewed cargo aircraft where the route is genuinely dangerous, where crew availability is the constraint, or where the mission profile is too long or dull for a crew — aerial refuelling being the clearest military case. For routine resupply where roads exist, compare honestly against ground transport, which is usually cheaper per tonne-kilometer by a wide margin. For contested resupply, weigh a small number of large aircraft against many small ones, since dispersal usually survives better than concentration.

Payload a few hundred kilograms on purpose-built designs to several tonnes on converted light transports · endurance several hours to about a day · cruise typically 150–200 knots on turboprop conversions · MQ-25 requirement 15,000 lb of fuel delivered at 500 nautical miles.

Cargo aircraft are large, slow, and non-stealthy, so any air defense engages them. Their routes are predictable because cargo goes to known places. Electronic attack on control links matters for approach and landing phases in particular. The practical constraint in most theatres is that a cargo aircraft flying to a forward position advertises that position, which is a planning problem more than a technical one.

Boeing MQ-25 Stingray (carrier-based uncrewed tanker), Reliable Robotics and Xwing autonomy conversions of Cessna Caravans, Elroy Air Chaparral, Sabrewing, Kaman KARGO UAV, and the various cargo conversion programs for existing light transports.

Commercial progress has been slower than forecast, largely because certification for uncrewed operation in civil airspace is a long process and because the crew-cost savings are smaller than the pitch decks assumed. The military side has done better: the MQ-25 addresses a specific problem — extending carrier air wing range without spending fighters on tanking — where the value is clear. Expect military and contested-logistics applications to lead, with civil freight following certification rather than the other way round.

2 minTesting MQ-25 Aboard an Aircraft CarrierBoeing · 100k+ views

2 minReliable Robotics remotely operates a large cargo aircraft with no one on boardReliable Robotics · 10k+ views

4 minXwing - Unveiling the World's First Autonomous Regional Cargo Aircraft (Modified Cessna 208B).Xwing · 5k+ views

Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service) · Unmanned Aircraft Systems (Federal Aviation Administration)

A man-portable loitering munition is a tube-launched drone carrying a small warhead, packed with its launcher into a rucksack weighing 3–10 kg. It flies for 15–40 minutes at ranges of 10–20 km, streams video to the operator, and dives onto a target when commanded. Some can wave off and re-attack. The warhead is 0.3–1.5 kg, sized for personnel, light vehicles, and crew-served weapons. Functionally it is a precision weapon that a soldier carries, with the guidance and the decision both provided by the operator watching the video.

It gives a small unit precision fires without calling for artillery or aircraft, and the man-in-the-loop lets the operator confirm the target and abort if something changes, which matters under restrictive rules of engagement. Weight is low enough to carry on foot. The weaknesses are cost per shot relative to an FPV drone doing a similar job, small warhead, short range, and vulnerability to jamming of the video and command links. Wind limits use, and the operator is occupied for the whole flight.

Use man-portable loitering munitions where a small unit needs precision effect against a specific target beyond the range of direct fire and cannot wait for supporting arms. If the target is inside a few kilometers and cost matters, an FPV drone does the same job for a tenth of the price, which is exactly why FPVs have displaced these weapons in some roles. If the target is armor, a larger anti-armor loitering munition or an anti-tank guided missile is the right choice — a 1 kg warhead will not defeat a main battle tank frontally.

System weight 3–10 kg including launcher · endurance 15–40 minutes · range 10–20 km · warhead 0.3–1.5 kg · unit cost $20k–80k.

Jamming the command and video link is the standard counter and is effective against most models. GNSS denial degrades navigation on models without visual terminal guidance. Small arms and shotguns work at short range. Because the operator must maintain the link, direction finding on that link can locate the launch position, which is a significant risk in a well-equipped opponent's area.

AeroVironment Switchblade 300, UVision Hero-30, WB Group Warmate, Rheinmetall and Elbit man-portable systems, and a growing number of Ukrainian, Turkish, and Chinese equivalents.

Unit costs run $20k–80k, which is far above an FPV drone and far below a guided missile. That middle position is under pressure from both directions: FPVs undercut them on cost for short-range work, and cheap fixed-wing one-way attack drones undercut them on range. The segment's durable advantage is packaging and reliability — a sealed tube that works when a soldier pulls it out of a rucksack after a month in the field, which improvised systems do not reliably deliver.

8 minUkraine Using Switchblade 300 'Kamikaze Drones' / Loitering MunitionsThe Armourer's Bench · 50k+ views

2 minAeroVironment Introduces Switchblade 300 Block 20 Rapidly Deployable Loitering Missile SystemGlobal Update · 5k+ views

2 minHero-30 Loitering munition system - Full missionUVision Air Ltd · 10k+ views

Defense Primer: Categories of Uncrewed Aircraft Systems (Congressional Research Service) · Loitering Munitions In Focus (Center for the Study of the Drone at Bard College)

An anti-armor loitering munition carries a shaped-charge or explosively formed penetrator warhead of 2–10 kg, loiters for 30–90 minutes at 40–80 km, and attacks from above where armor is thinnest. Launch is from a tube on a vehicle or a two-person team. Guidance is operator-directed through a video link for most of the flight, with terminal lock-on that completes the attack if the link is lost. The combination of loiter time and top attack makes it effective against targets that are hidden, moving, or dispersed, which artillery handles poorly.

The ability to search for a target, wait, and then attack the roof of an armored vehicle is a genuinely different capability from a direct-fire missile. Range of 40–80 km puts it well beyond most direct-fire and short-range indirect systems. The weaknesses are cost, which sits in the same range as anti-tank guided missiles without the same reliability against a maneuvering target under active protection, and dependence on the operator link for target selection. Flight time also means the target may move or be reinforced during the engagement.

Use anti-armor loitering munitions against armored vehicles that are dispersed, concealed, or beyond the range of direct-fire weapons, and against high-value systems like air-defense radars and artillery. If the target is in view and inside 4 km, an anti-tank guided missile is faster and more certain. If the target is soft and inside 10 km, an FPV drone costs a fraction as much. The best use has generally been against high-value targets where the loiter time is what makes the engagement possible at all.

Warhead 2–10 kg shaped charge or explosively formed penetrator · loiter 30–90 minutes · range 40–80 km · unit cost $100k–300k.

Active protection systems on vehicles engage top-attack munitions, though performance against slow-diving drones varies. Jamming the operator link is effective against models that need it for target selection. Cage and slat armor over vehicle roofs, now widely improvised, defeats some shaped charges. Concealment and dispersion reduce the number of targets worth the munition's cost, which is itself a form of counter.

AeroVironment Switchblade 600, UVision Hero-120 and Hero-400, Russian Lancet-3 and its derivatives, Israeli Spike Firefly, and the Turkish Kargu and Alpagu family.

Unit costs of $100k–300k put these weapons in direct competition with anti-tank guided missiles, and the choice usually comes down to whether loiter time is worth the price. The Lancet's reported effectiveness against Ukrainian artillery, at a unit cost well below the systems it destroyed, is the standard case study for favorable cost exchange. Production rate rather than unit performance has been the constraint on both sides, and it remains the key metric for any program in this class.

5 minSwitchblade 600 Loitering missile | Kamikazi drone How it works #loiteringmunition #dronesAiTelly · 500k+ views

3 minSwitchblade 600 Loitering MunitionAV · 10k+ views

2 minHERO 120 Loitering Munition - UVision airUVision Air Ltd · 10k+ views

Defense Primer: Categories of Uncrewed Aircraft Systems (Congressional Research Service) · The Uninhabited War in Ukraine (International Institute for Strategic Studies)

A long-range one-way attack drone is a small propeller-driven aircraft carrying a 20–90 kg warhead to targets 800–2,500 km away. Navigation is inertial plus GNSS, sometimes with terminal visual or scene matching, and there is no return trip and usually no operator link. The Shahed-136 established the archetype: a delta-wing airframe with a small piston or expendable turbojet engine, built for a reported $20k–50k per unit, launched in salvos from a simple rail. It is a cruise missile built to consumer-electronics cost standards rather than to missile standards.

Cost per kilometer of range is the whole point. These aircraft deliver a warhead a thousand kilometers for the price of a car, and salvos of dozens saturate air defenses whose interceptors cost ten to a hundred times more per shot. Production is simple enough to scale quickly with a non-aerospace supply chain. The weaknesses are speed and accuracy: 100–200 km/h makes them engageable by almost anything, and accuracy of tens of meters suits area targets rather than point targets. Payload is small relative to a real cruise missile.

Use long-range one-way attack drones for sustained pressure on fixed infrastructure and area targets, and for saturating air defenses ahead of more capable weapons. If the target is hardened or requires precision, a cruise or ballistic missile is necessary and costs accordingly. If the target is within 30 km, cheaper short-range systems apply. The doctrinal significance is the cost exchange: forcing a defender to spend $1M interceptors on $30k drones is a losing proposition for the defender even when every shot hits, which is why cheap interceptors and guns have come back into focus.

Range 800–2,500 km · warhead 20–90 kg · cruise speed 100–200 km/h · accuracy tens of meters · unit cost $20k–50k on reported figures · defensive interceptors $0.5–4M per shot.

Guns, cheap interceptors, and electronic warfare all work, and the whole counter-UAS field has reoriented around making the cost exchange favorable. GNSS jamming degrades navigation, though terminal visual guidance and inertial coasting have reduced its effectiveness. They are slow and their acoustic signature is distinctive, which makes distributed detection networks practical. The real problem for defenders is volume rather than difficulty per engagement.

Shahed-136 and its Russian-produced Geran-2 variant, Ukrainian long-range strike drones including the Liutyi, the Iranian Shahed-131 and 238 jet variant, and several emerging Western attempts to field comparable low-cost strike drones.

This class has done more to change strike economics than any other recent system. Reported unit costs of $20k–50k against defensive interceptors at $0.5–4M each produce an exchange ratio no defender can sustain, and production at thousands per month from workshop-scale facilities is achievable. Western programs are now explicitly chasing the same cost point, and the difficulty they face is not technical but industrial: building a $30k airframe requires a supply chain and a quality standard that traditional defense manufacturing does not have.

7 minKamikaze drone Iran Shahed 136 | How it WorksAiTelly · 5m+ views

14 minThe $30K Drone That Forces $40M Missiles To Launch: Shahed-136 and Geran-2SYG DESIGNWORKS · 10k+ views

9 minSHAHED 136 drone engine TEAR DOWNUA COURAGE · 100k+ views

Iran's Transfer of Weaponry to Russia for Use in Ukraine (Congressional Research Service) · Shahed-131 and -136 (CSIS Missile Defense Project)

An anti-radiation loitering munition carries a passive radar seeker and orbits a defended area waiting for an air-defense radar to transmit. When one does, it homes on the emission and attacks. This inverts the timing problem that anti-radiation missiles have: a missile launched at a radar can be defeated by the radar shutting down, but a drone that can loiter for two to nine hours simply waits, and the radar operator has to stay off the air for the whole period. The Harpy pioneered the concept in the 1990s and it has since been widely copied.

Patience is the capability. It converts an opponent's air defense from a system that can be temporarily suppressed into one that must stay silent for hours, which is often operationally equivalent to being destroyed. Fully autonomous variants need no operator link once launched. The weaknesses are that decoy emitters draw them cheaply, discriminating the intended radar in a dense electromagnetic environment is hard, and full autonomy in target selection raises legal and policy questions that some countries treat as disqualifying.

Use anti-radiation loitering munitions in suppression and destruction of enemy air defenses, particularly at the start of a campaign when radars must operate and cannot simply hide. If the radar's location is known and it will be transmitting at a predictable time, a conventional anti-radiation missile is faster and cheaper per shot. If the target set is broader than radars, a general loitering munition with an imaging seeker is more flexible. Expect decoy emitters, and plan the salvo size on the assumption that some munitions will attack decoys.

Loiter endurance 2–9 hours · range 200–1,000 km on fielded types · launch weight roughly 100–200 kg · warhead 15–35 kg · unit cost $100k–1M.

Cheap decoy emitters are the standard and effective counter, and they cost a small fraction of the munition. Emission control works if the defender can accept being blind for the loiter period. Point defense guns and short-range missiles engage them since they are slow. Some newer air-defense systems specifically pair a remote illuminating decoy with the real radar so that any attack goes to the decoy.

IAI Harpy and Harop, the Turkish Kargu and Alpagu in related roles, the Chinese ASN-301 (a Harpy derivative), and the anti-radiation modes on several general-purpose loitering munitions.

Unit costs of $100k–1M sit below anti-radiation missiles while offering loiter time that missiles cannot. The export market has been active, with Harop sales to several countries and demonstrated use in the Nagorno-Karabakh conflict, where the combination of loitering munitions and armed drones against unprepared air defenses was decisive. The counter-economics — cheap decoys against expensive munitions — is the segment's structural weakness and is well understood by defenders.

4 minLoitering Munition 9-Hour Endurance Strike with Precision Targeting | HAROP - IAIIAI · 50k+ views

5 minHow the Israel's Harop suicide Drone Changed Modern WarfareThe Aviation Archives · 5k+ views

2 minMaritime Harop - Loitering Munition SystemIAI · 10k+ views

Defense Primer: Categories of Uncrewed Aircraft Systems (Congressional Research Service) · Mapping the Development of Autonomy in Weapon Systems (Stockholm International Peace Research Institute)

A swarm munition system launches many small drones from a canister or a tube, either from an aircraft, a vehicle, or a ship. The individual aircraft are small and cheap, carry a small warhead or a sensor, and coordinate with each other over a mesh network — sharing detections, dividing a target area, and reallocating when members are lost. The point is not any one aircraft's capability but the collective behavior: dozens of objects arriving from multiple directions saturate defenses that could handle any one of them easily.

Saturation and graceful degradation are the strengths. A defense that can engage four targets per minute is overwhelmed by forty arriving together, and losing half a swarm degrades its output rather than ending the mission. Cost per aircraft is low. The weaknesses are the coordination layer and the policy questions. Mesh networks are jammable, autonomous target selection within a swarm is legally contested, and demonstrating swarm behavior in a test range is much easier than making it work in an environment with jamming, terrain, and unexpected targets. Payload per aircraft is very small.

Consider swarm munitions for saturating point defenses, for wide-area search where many cheap sensors beat one good one, and for distributed electronic attack. If the target requires a substantial warhead, a small swarm aircraft cannot carry it and a conventional munition is necessary. If the environment is heavily jammed, the coordination that makes a swarm more than a group of individual drones may not survive, so check what the system does when the mesh degrades — the honest answer is usually that it becomes a set of independent waypoint-following drones.

Per-aircraft mass roughly 0.3–7 kg · endurance minutes to about an hour · range tens of kilometers from the launch point · payload a few kilograms at most · demonstrated swarm sizes 30–100 aircraft.

Jamming the mesh network reduces a swarm to uncoordinated individuals. High-power microwave weapons are the most promising kinetic counter, because they engage many targets in one shot, which is the only way to match the cost exchange. Guns with programmable airburst ammunition and short-range interceptors both work per target but lose on economics against large numbers. Distributed detection is straightforward given the acoustic and RF signatures.

Raytheon Coyote in swarm configurations, the US Navy LOCUST and DARPA OFFSET and Gremlins programs, the Perdix micro-drone demonstration launched from fighter dispensers, the Turkish Kargu swarm demonstrations, and Chinese canister-launched swarm systems shown at exhibitions.

Swarms are heavily funded on the argument that they invert the cost exchange in the attacker's favor, and the counter-investment in high-power microwave weapons follows the same logic in reverse. Progress has been slower than demonstrations suggest, because the difficult part is robust autonomy in degraded conditions rather than the coordination algorithms shown in clear-weather tests. The systems that field first are likely to be simple ones with modest coordination rather than the fully autonomous swarms of the concept videos.

4 minPerdix Drone Swarm – Fighters Release Hive-mind-controlled Weapon UAVs in AirAiirSource Military · 100k+ views

2 minWatch the Navy's LOCUST launcher fire a swarm of dronesBusiness Insider · 1m+ views

5 minDrone Swarms, ExplainedDEFENCE CENTRAL · 10k+ views

Emerging Military Technologies: Background and Issues for Congress (Congressional Research Service) · Research programs (DARPA)

A small attack uncrewed surface vessel is a fast boat, typically 5–8 m long, carrying several hundred kilograms of explosive and steered by an operator over a satellite or radio link with video from an onboard camera. Speed is 40–80 km/h and range is 400–800 km. Ukraine developed the type into an operationally decisive weapon against the Black Sea Fleet, using boats built from commercial hulls, marine engines, Starlink terminals, and off-the-shelf cameras at a cost of $200k–500k each against warships worth hundreds of millions.

The cost exchange is extraordinary and it works: a $250k boat that damages or sinks a frigate is a trade no navy can accept repeatedly. The vessels are small, sit low in the water, and are hard to detect on surface radar in any sea state, especially at night. They can be built in dispersed workshops. The weaknesses are sea state, which limits operations above roughly sea state 3–4, dependence on a control link for terminal guidance on most designs, and the fact that a warship alert to the threat with functioning close-in defenses can engage them.

Use small attack USVs against high-value naval targets and port infrastructure when you lack anti-ship missiles or want to complicate a defender's problem by attacking from the surface as well as the air. If the target is at sea and defended with layered air and surface defenses, expect to need salvos and to combine them with aerial attack. For a defender, the relevant conclusion is that harbours and anchorages are no longer safe from a technically unsophisticated opponent, and physical barriers plus persistent surveillance matter more than they did.

Hull length 5–8 m · warhead several hundred kilograms · speed 40–80 km/h · range 400–800 km · usable up to about sea state 3–4 · unit cost $200k–500k.

Helicopter-mounted guns and small arms are effective when the boats are detected in time, which is the hard part. Close-in weapon systems engage them but were designed for aircraft and missiles rather than small surface targets. Jamming the control link, including satellite terminals, degrades terminal guidance. Physical barriers — booms, nets, and blockships — at harbour entrances are the most reliable defense and have been widely reinstated.

Ukrainian Magura V5 and Sea Baby, which between them have damaged or destroyed a significant portion of the Russian Black Sea Fleet, Houthi explosive boats used against shipping in the Red Sea, and a growing number of Western and Asian programs developing comparable craft.

This is the clearest recent case of an asymmetric capability built from commercial components: hulls, outboard or inboard marine engines, satellite terminals, and cameras, integrated by small teams for a few hundred thousand dollars. The strategic effect has been out of all proportion to the investment. Every navy is now reassessing harbour defense and close-in surface engagement, which is a much larger expenditure than the boats that prompted it.

4 minUkraine's kamikaze drone Magura V5 in detailsNaval News · 10k+ views

8 minNaval Kamikaze Sea Drone How it works using Starlink SatelliteAiTelly · 1m+ views

4 minUKRAINE’S MAGURA NAVAL DRONES: BLACK SEA EQUALIZERSU.S. Naval Institute · 5k+ views

Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service) · Uncrewed Platforms Have Been Critical to Ukraine's Success in the Black Sea (Royal United Services Institute)

A medium uncrewed surface vessel is a 12–50 m ship displacing 100–500 tonnes, designed to operate for 30–60 days with no crew aboard and only periodic supervision from ashore. It carries sensors — surface search radar, electro-optical systems, towed sonar arrays — rather than weapons, and its purpose is to add sensing mass to a fleet at a cost per hull that allows numbers. Sea Hunter demonstrated autonomous transits from San Diego to Hawaii, including compliance with the international rules for preventing collisions at sea.

Persistence and cost per hull are the strengths. A vessel that stays out for 60 days without a crew rotation gives sensor coverage that a crewed ship cannot match per dollar, and it can be sent into areas where risking a crew would be unacceptable. The weaknesses are reliability and law. There is no crew to fix a failed pump or a fouled propeller, so mean time between failures has to be far better than a crewed ship's, and it currently is not. Legal status under maritime law is unsettled, and collision avoidance in busy waters remains a source of risk.

Use medium USVs to extend a fleet's sensor picture over wide ocean areas, for persistent presence in contested waters, and for anti-submarine trailing where a cheap platform can follow a submarine that a destroyer cannot afford to shadow. If the mission needs weapons employment decisions or boarding, a crewed vessel is required. If the area is small and near shore, smaller uncrewed craft or aircraft are cheaper. Treat reliability engineering as the program's central challenge — the record of these programs is dominated by maintenance failures rather than by autonomy failures.

Length 12–50 m · displacement 100–500 tonnes · endurance 30–60 days unattended · cruise speed roughly 10–12 knots · unit cost $20–50M per hull.

An undefended 200-tonne ship is easily attacked by anything from a missile to a boarding party, and several navies have noted that an uncrewed vessel with no crew aboard is also an attractive target for capture. Jamming satellite links limits supervision, though the vessels are designed to continue autonomously. The most common defeat has been mechanical: propulsion and generator failures ending deployments early.

DARPA and US Navy Sea Hunter and Seahawk, the Overlord USV program vessels Nomad and Ranger, Saildrone Voyager and Surveyor in surveillance roles, and Chinese and Israeli medium USV programs.

US Navy programs in this class have had a difficult history, with repeated restructuring driven by reliability shortfalls and by uncertainty about what the vessels should actually do. Cost per hull of $20–50M is far below a crewed combatant but far above the small craft that have proven effective. The most commercially successful vessels in the category have been the commercial ocean-data platforms, where the business model is selling data rather than selling ships.

3 minACTUV Construction Timelapse and WalkthroughDARPAtv · 10k+ views

6 minU.S. NAVY’S SEA HUNTER UNMANNED SURFACE VESSEL - FULL ANALYSISDefense Updates · 10k+ views

9 minUS Navy Unmanned Surface Vessel Division OneNaval News · 50k+ views

Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service)

A large uncrewed surface vessel is a 60–90 m ship of 1,000–2,000 tonnes intended to carry vertical launch cells as distributed missile magazines for a fleet. The concept separates the magazine from the expensive combat system: a destroyer with a good radar and combat management system directs missiles launched from cheaper uncrewed hulls sailing nearby, which multiplies fleet missile capacity without multiplying billion-dollar warships. Designs are usually based on commercial offshore supply vessel hulls to keep cost and construction risk down.

Magazine depth at lower cost per cell is the argument, and fleet missile capacity is a genuine constraint in any serious naval conflict. Commercial hull forms and machinery reduce build cost and time. The weaknesses are substantial. Reliability without a crew over a 60-day deployment has not been demonstrated at this scale. Weapons release from an uncrewed platform raises command and control questions that navies are still working through. And there is a persistent argument that a lightly defended missile barge is simply an attractive target that must be escorted, which erodes the cost advantage.

The case for large USVs is strongest where fleet magazine depth is the binding constraint and where the vessels can operate under the protection of crewed combatants. If the requirement is sensing rather than magazine capacity, a medium USV or an aircraft is far cheaper. If survivability in a contested environment matters, an escorted uncrewed barge may not be cheaper than another crewed combatant once escort costs are counted. Programs in this class should be judged on demonstrated unattended reliability more than on cost per missile cell.

Length 60–90 m · displacement 1,000–2,000 tonnes · 16–32 vertical launch cells in the US Navy requirement · endurance targeted at about 60 days unattended · unit cost roughly $250–500M.

An uncrewed vessel with no self-defense is engageable by anti-ship missiles, submarines, and small craft. Capture is a real concern, since a vessel with no crew cannot resist boarding and carries valuable weapons and equipment. Disrupting the command link between the controlling warship and the launcher degrades or prevents weapons employment, which is a specific vulnerability of the distributed magazine concept.

The US Navy's Large Unmanned Surface Vessel program and its predecessor Overlord vessels, the Ghost Fleet Overlord conversions Nomad and Ranger which conducted long autonomous transits, and Chinese programs including large uncrewed vessels shown in recent years.

The LUSV program has been repeatedly delayed and restructured by Congress and the Navy, with concerns about reliability, requirements, and whether the concept is sound. Costs of roughly $250–500M per vessel have been discussed, which is well below a destroyer but high enough that losing one matters. The unresolved question is whether the fleet wants many cheap magazines or fewer capable ships, and that is a doctrinal argument rather than a technical one.

12 minU.S NAVY GHOST FLEET WILL BE HAVE LARGE UNMANNED SURFACE VESSELS (LUSV) !!Defense Updates · 50k+ views

9 minGhost Fleet OverlordDark Tech · 100k+ views

Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service)

A long-endurance survey USV uses wind, wave, or solar energy for propulsion, so it stays at sea for months rather than days. Sail-powered designs carry a rigid wing and generate electrical power from solar panels; wave gliders extract propulsion directly from wave motion through a submerged articulated system. Speeds are slow — 1–5 knots — and payloads are modest, but the vessels operate for 6–12 months per deployment at a fuel cost of zero, which changes what persistent ocean observation costs by roughly two orders of magnitude.

Persistence is unmatched: a year at sea with no refuelling and no crew. Operating cost is dominated by satellite data charges rather than by fuel or personnel. The vessels are small and unremarkable, which suits covert observation. The weaknesses are speed and payload. At 3 knots, a vessel cannot be repositioned quickly or hold station against a strong current, and payload power budgets of tens to a few hundred watts constrain what sensors can run. Fouling, collisions with shipping, and fishing-gear entanglement account for most losses.

Use long-endurance USVs for persistent ocean observation, maritime domain awareness in wide areas, acoustic monitoring, and hydrographic and environmental survey. If you need to respond to an event or reposition quickly, a powered vessel or an aircraft is necessary. If the sensing requirement needs kilowatts, the power budget will not support it. The commercial model — selling the data rather than the vessel — has worked well here, and buyers should compare a data subscription against owning and operating hulls.

Speed 1–5 knots · deployment 6–12 months per launch · hull length roughly 3–20 m across fielded types · payload power budget tens to a few hundred watts · zero fuel consumption, so operating cost is mostly satellite data charges.

These vessels are slow and defenseless, so anything that finds one can destroy or capture it, and several have been seized. Their small size makes detection hard, which is their main protection. Fouling and biological growth degrade performance over long deployments. Because they cannot outrun weather, severe storms are a real loss mechanism.

Saildrone Explorer, Voyager, and Surveyor, Liquid Robotics Wave Glider, Ocius Bluebottle, AutoNaut wave-propelled vessels, and the growing use of these platforms by NOAA, navies, and offshore energy operators.

This is one of the few uncrewed maritime segments with a working commercial market, driven by ocean science, fisheries enforcement, offshore energy survey, and maritime domain awareness. Data-as-a-service pricing has been the successful model. Defense interest has grown quickly because persistent presence in wide ocean areas at low cost is exactly what maritime awareness requires, and because the same hull can carry acoustic sensors for undersea monitoring.

2 minThe Wave Glider: How it Works (with audio)Liquid Robotics · 100k+ views

5 minHow a fleet of sailing drones is monitoring our oceansSaildrone · 10k+ views

4 minSaildrones: Cutting Edge Technology for Ocean ResearchNOAA Fisheries · 50k+ views

Uncrewed systems overview (NOAA Office of Marine and Aviation Operations) · Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service)

A man-portable uncrewed underwater vehicle weighs 10–50 kg, is launched over the side by one or two people, and runs preprogrammed survey patterns for 8–20 hours. Payload is usually a side-scan or synthetic aperture sonar for mine hunting and seabed survey, plus a camera for close inspection. Navigation underwater is inertial aided by a Doppler velocity log, with periodic surfacing for a GNSS fix. There is essentially no communication while submerged beyond low-rate acoustic links, so the vehicle must complete its mission on its own and report when it comes up.

It removes divers from mine countermeasures, which is the single most dangerous routine task in naval operations, and it does so from small boats without specialized ships. Cost of $100k–500k per vehicle allows navies to buy them in numbers. The weaknesses are endurance, coverage rate, and communications. Twenty hours at 3 knots covers a modest area, so clearing a large minefield takes many vehicle-days. Underwater communication is slow and unreliable, so there is no supervision during the mission, and a vehicle that fails is usually lost.

Use man-portable UUVs for mine countermeasures, harbour and hull inspection, and seabed survey in shallow water where a diver would otherwise go. If the area is large or deep, a larger vehicle with better endurance and sonar range covers ground far faster. If the requirement is real-time inspection with manipulation, a tethered remotely operated vehicle gives control and power that an autonomous vehicle cannot. Plan coverage rate rather than endurance — the useful metric is square kilometers per vehicle-day at the required detection confidence.

Mass 10–50 kg · endurance 8–20 hours · survey speed around 3 knots · depth rating typically 100–300 m · unit cost $100k–500k.

Underwater vehicles are hard to detect and hard to attack, which is their principal protection. Countermeasures are mostly environmental: strong currents, poor visibility, cluttered seabeds that hide mines among debris, and shallow water that limits sonar geometry. Acoustic communication jamming is possible but rarely relevant. The realistic loss mechanism is vehicle failure or navigation error rather than enemy action.

Teledyne Gavia and the REMUS family (REMUS 100 and 300) from HII, Saab AUV62, Kongsberg HUGIN in its smaller variants, ECA Group A9 and A18, and the very large number of these vehicles in mine countermeasures service worldwide.

This is the most mature and commercially healthy uncrewed underwater segment, with a real market spanning defense mine countermeasures, offshore energy survey, and ocean science. Prices have fallen and capability has risen steadily. Navies are replacing crewed mine countermeasures vessels with uncrewed systems deployed from ordinary ships, which is a substantial procurement shift and the main growth driver for the category.

3 minREMUS 100: The Industry Standard Compact Man-Portable AUVHydroid, Inc. · 5k+ views

3 minREMUS 100 AnimationHydroid, Inc. · 1k+ views

3 minTeledyne Gavia Osprey Autonomous Underwater VehicleTeledyne Marine · 10k+ views

Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service) · Autonomous underwater vehicles (Woods Hole Oceanographic Institution)

A medium UUV is a torpedo-shaped vehicle 3–7 m long, launched from a ship's crane or a submarine's torpedo tube, running for two to five days at depths to 3,000–6,000 m. Payloads include synthetic aperture sonar, multibeam echosounders, sub-bottom profilers, magnetometers, and environmental sensors. Navigation is high-grade inertial aided by Doppler velocity log, with error growth of a fraction of a percent of distance travelled, and increasingly with terrain-referenced navigation against previously mapped seabed.

Area coverage and depth capability are the strengths: these vehicles map the deep seabed at resolutions that would require far more expensive ship time to achieve by towed sonar. Multi-day endurance means a single deployment covers a large survey block. The weaknesses are cost, handling, and recovery. Launch and recovery in a seaway is difficult and is where vehicles get damaged, deep-rated pressure housings are expensive, and a vehicle lost at 4,000 m is not coming back. Data cannot be reviewed until recovery, so a sensor misconfiguration wastes the whole mission.

Use medium UUVs for deep-water survey, seabed infrastructure inspection, wide-area mine reconnaissance, and intelligence preparation of the undersea environment. If the water is shallow and the area small, a man-portable vehicle is cheaper and easier to handle. If the task needs manipulation or live decision-making, a tethered vehicle is the answer. For infrastructure monitoring, compare against permanent seabed sensors and against fiber-optic distributed acoustic sensing on the cable itself, which can be much cheaper for continuous monitoring.

Length 3–7 m · endurance 2–5 days · depth rating 3,000–6,000 m · survey speed roughly 3–4 knots · navigation drift a fraction of a percent of distance travelled · unit cost $2–10M.

Detection of a quiet UUV at depth is genuinely difficult, which is why they are attractive for intelligence work. The vulnerabilities are operational: launch and recovery must happen from a ship that is visible, communications are limited to acoustic links and surface periods, and navigation error accumulates without external fixes. Capture of a vehicle that surfaces in the wrong place has happened and is a real intelligence loss.

Kongsberg HUGIN, Teledyne Gavia and Slocum-derived designs, Bluefin-21 (used in the MH370 search), Boeing Echo Ranger, and the survey fleets operated by offshore energy contractors and hydrographic agencies.

The offshore energy and hydrographic survey markets fund most development in this class, and defense buys into a supply base that commerce sustains. Vehicle costs of $2–10M are substantial but compare well against the ship time they replace. Growth drivers now include undersea infrastructure protection, which has become a policy priority after several cable and pipeline incidents, and which needs exactly the survey and monitoring capability these vehicles provide.

6 minExploring Earth’s Final Frontier With AUV TechnologyFreethink · 100k+ views

4 minKongsberg Maritime's autonomous underwater vehicle "Hugin"Simrad by Kongsberg · 5k+ views

5 minDeep Sea Survey and Mapping with the Teledyne Gavia SeaRaptor AUVTeledyne Marine Vehicles · 10k+ views

Underwater vehicles (Woods Hole Oceanographic Institution) · Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service)

A large displacement UUV is 8–15 m long and displaces several tonnes, too large for a torpedo tube, so it launches from a ship, a pier, or a dedicated dry deck shelter. Endurance runs 30–70 days on lithium batteries or fuel cells, and payload volume allows a substantial sensor fit or deployable subsystems. The design intent is to do submarine-like intelligence and survey work in places where risking a crewed submarine is unwise, at perhaps a hundredth of the cost per day at sea.

Persistence in denied areas without risking a crew is the argument, and it is a strong one for intelligence preparation, seabed survey, and acoustic monitoring near an opponent's coast. Cost per day at sea is dramatically below a crewed submarine. The weaknesses are energy and reliability. Battery energy density caps endurance and speed, so these vehicles move slowly and cannot reposition quickly; fuel cells help but add complexity. And 60 days of unattended operation demands a reliability standard that has been hard to reach, with several programs delayed by exactly this.

Use large UUVs for persistent undersea intelligence, seabed mapping and infrastructure survey in sensitive areas, and as deployable sensor carriers. If the mission needs speed or weapons employment, a crewed submarine or a torpedo is the answer. If it needs only a survey of a defined block, a medium UUV launched from a ship is cheaper. The pacing question for any program here is demonstrated unattended reliability over the full mission duration, which is where nearly all schedule slip has occurred.

Length 8–15 m · displacement several tonnes · endurance 30–70 days on batteries or fuel cells · transit speed a few knots · cost per day at sea roughly 1% of a crewed submarine's.

Detection is difficult, which is the vehicle's main protection. Practical vulnerabilities are the launch platform, the surfacing events needed for communication and navigation fixes, and mechanical failure. As with all uncrewed vehicles, one that is found can be recovered by an opponent, and the intelligence loss from a captured vehicle with its mission data can exceed the value of the vehicle itself.

Boeing Orca XLUUV in its smaller configurations, the Snakehead LDUUV program, Anduril Dive-LD and Dive-XL, Saab and Kongsberg large vehicle programs, and Chinese large UUV designs shown publicly in recent years.

Programs have generally cost more and taken longer than planned. The commercial entrants — Anduril in particular — argue that building to commercial reliability standards and iterating quickly beats the traditional development approach, and their vehicles have reached the water faster than several government programs. Whether that holds through long unattended deployments is the open question and it will be answered by operational experience rather than by testing.

8 min#USNavy has christened its largest Unmanned Undersea Vehicle!Defense Updates · 10k+ views

17 minThe Underwater Drone That Latches Onto Ships And Launches UAVsTask & Purpose · 100k+ views

Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service) · Underwater vehicles (Woods Hole Oceanographic Institution)

An extra-large UUV displaces 50–80 tonnes and is 20–25 m long — a small submarine without a crew. It launches from a pier rather than a ship, transits thousands of kilometers on diesel-electric or fuel-cell power, and carries a modular payload bay for mines, sensor packages, smaller vehicles, or effectors. Orca, the leading Western example, is designed for a payload section that can be reconfigured for different missions, and mine laying in denied waters has been the first stated operational use.

Range and payload volume approach useful submarine-like capability at a small fraction of a submarine's cost and with no crew at risk. Autonomous mine laying in an opponent's approaches is a genuinely valuable mission that a crewed submarine can do but at much higher risk. The weaknesses are program risk and command. These vehicles are complex enough to inherit shipbuilding-scale problems, the Orca program ran years late and well over budget, and questions about employing weapons from an unattended autonomous platform remain unsettled in policy terms.

The clearest case is mine laying and payload delivery in areas where sending a crewed submarine would be unacceptable, and long-duration sensing in denied waters. If the mission is survey or inspection, smaller vehicles do it far more cheaply. If it requires weapons employment against moving targets, the autonomy and command questions are not resolved and a crewed platform is appropriate. Programs in this class should be assessed on delivered vehicles at sea rather than on capability projections, given the sector's record.

Displacement 50–80 tonnes · length 20–25 m · range several thousand nautical miles on diesel-electric or fuel-cell power · endurance weeks to months · modular payload bay for mines, sensors, or smaller vehicles · unit cost in the tens of millions of dollars.

Anti-submarine warfare techniques apply, though a small quiet vehicle is a harder acoustic target than a crewed submarine. The base and the launch point are visible and targetable. Long transits at low speed mean an opponent with good undersea surveillance has time to react. Capture of a vehicle carrying weapons or mission data would be a significant intelligence event, and the vehicle cannot resist.

Boeing Orca XLUUV for the US Navy, Anduril's Ghost Shark developed with Australia, the UK's Cetus demonstrator, and reported Chinese and Russian extra-large vehicle programs including the nuclear-powered Poseidon, which is a different concept but occupies related conceptual space.

Orca is the cautionary example: a program intended to be a fast, low-cost acquisition that ran substantially late and over budget, largely because building a 50-tonne autonomous submarine turned out to be a shipbuilding project rather than a drone project. Australia's Ghost Shark program has moved faster with a commercial development model. Unit costs in the tens of millions are far below a submarine but high enough that these are assets to be protected, not attrited.

7 minORCA XLUUV - Boeing has delivered the first one to the #USNavy !Defense Updates · 100k+ views

3 minAnduril Opens Ghost Shark Factory as First Autonomous Underwater Vehicle Rolls Off the LineIEN Magazine · 1k+ views

Navy Large Unmanned Surface Vessels: Background and Issues for Congress (Congressional Research Service)

An underwater glider has no propeller. It changes its buoyancy with a small pump, sinks and rises through the water column, and converts that vertical motion into forward travel with fixed wings, producing a sawtooth path through the ocean. Forward speed is about 0.5 knots, but power consumption is a fraction of a watt on average, so a glider operates for six months to a year and covers thousands of kilometers on a battery pack. It surfaces periodically to report data and receive new instructions over satellite.

Endurance per unit of energy is extraordinary and unmatched by any powered vehicle. Gliders are quiet — there is no propeller noise — which makes them excellent acoustic sensing platforms. Cost of $100k–200k allows fleets. The weaknesses are speed and payload. At half a knot, a glider cannot hold station against a current or respond to events, and the power budget supports only low-power sensors: conductivity, temperature, depth, and passive acoustics rather than active sonar. Fishing gear and ship strikes account for most losses.

Use gliders for long-duration ocean sensing where slow is acceptable: oceanographic survey, acoustic monitoring for submarines, environmental characterization ahead of operations, and persistent presence in wide areas. If you need to respond to an event or survey a specific block quickly, a powered vehicle is necessary. Gliders work best deployed in fleets covering an area statistically rather than as individual assets tasked to specific points, and that is how oceanographic programs use them.

Forward speed about 0.5 knots · endurance 6–12 months · range thousands of kilometers per battery pack · average power draw a fraction of a watt · dive depth 200–1,000 m depending on the model · unit cost $100k–200k.

They are very hard to detect, and their main vulnerabilities are environmental. Recovery of a glider by a curious opponent has happened, notably the 2016 Chinese seizure of a US Navy glider in the South China Sea, which illustrated both the vulnerability and the political sensitivity. Strong currents can carry them off station. Biofouling degrades performance over long deployments.

Teledyne Slocum glider, Kongsberg Seaglider, the Spray glider, the US Navy's Littoral Battlespace Sensing gliders, and the large academic fleets operated by oceanographic institutions worldwide.

Gliders have a healthy commercial and scientific market and a mature supplier base, which keeps costs moderate and reliability good. Defense use rides on that base. The interesting development is acoustic payloads: a fleet of quiet gliders carrying hydrophones offers wide-area undersea surveillance at a small fraction of the cost of towed arrays or fixed installations, and that argument is driving renewed naval interest in a technology that oceanographers have used routinely for two decades.

5 minHow a glider works?SOCIB · 50k+ views

4 minHow Underwater Gliders Work Pt. 1NASLab · 10k+ views

3 minUnderwater Glider Silently Surveys the SeasPurdue University Mechanical Engineering · 10k+ views

Underwater vehicles (Woods Hole Oceanographic Institution) · Uncrewed systems overview (NOAA Office of Marine and Aviation Operations)

Small ground robots split into two related types. Throwable robots weigh 2–8 kg, survive being thrown through a window or down a stairwell, and carry cameras and microphones for building reconnaissance. Explosive ordnance disposal robots weigh 15–350 kg, drive on tracks, and carry a manipulator arm, cameras, and disruptors for examining and defeating suspected devices. Both are directly controlled by an operator over a radio link with video, and neither needs much autonomy because a human is watching continuously and the tasks require judgment.

They put a machine where a person would otherwise be, in the two situations where that is most obviously worth doing: entering an unknown building and approaching a possible bomb. Cost is low relative to the risk avoided, training is straightforward, and the technology is mature after two decades of counter-IED use. The weaknesses are mobility and communications. Stairs, rubble, mud, and soft ground defeat tracked robots regularly, and radio links fail inside buildings and underground, which is exactly where these robots are used.

Use throwable robots for building and tunnel reconnaissance ahead of people, and EOD robots for any approach to a suspected explosive device. If the environment blocks radio, use a tethered robot or one with a fiber spool, which is common in EOD work for exactly this reason. If the terrain is broken enough that a tracked chassis will not manage it, a legged robot may get there, at several times the cost. For large-area searching, aerial systems cover ground far faster.

Throwable robots 2–8 kg · EOD robots 15–350 kg · endurance a few hours per battery · radio control range a few hundred meters line of sight, much less inside structures · unit cost $10k for a throwable up to $250k for a large EOD system.

Radio links are the main vulnerability and they fail without any enemy action inside buildings. Deliberate jamming works. Physical obstacles — stairs, debris, doors — stop small robots routinely, and a defender who anticipates their use can arrange terrain to do so. Their approach is also conspicuous, so any element of surprise is lost.

Teledyne FLIR PackBot and FirstLook, the Recon Robotics Throwbot, Endeavor Robotics and QinetiQ Talon systems, the Israeli Roboteam Micro Tactical Ground Robot, and the very large numbers of EOD robots fielded during two decades of counter-IED operations.

This is a mature market with well-established suppliers, prices from $10k for a throwable robot to $250k for a large EOD system, and demand tied to counter-IED and police budgets. Growth has flattened as counterinsurgency operations wound down. The technology base has largely been absorbed into wider ground-robotics work, and the interesting developments now are in autonomy and in legged platforms rather than in the tracked chassis that defined the category.

3 minPackBot® 510 | Unmanned Ground SystemsTeledyne FLIR · 10k+ views

2 minA Robot You Can ThrowIEEE Spectrum · 10k+ views

4 minHow Do Robots Dismantle Bombs?Seeker · 50k+ views

The Army's Robotic Combat Vehicle (RCV) Program (Congressional Research Service)

A squad logistics UGV is a small tracked or wheeled vehicle carrying 200–500 kg of supplies, batteries, or casualties, following a squad on foot or driving a preplanned route. Speeds are 5–15 km/h and endurance is 60–100 km per charge or tank. The main autonomy mode is leader-follow, where the vehicle tracks a person or another vehicle, because full autonomous navigation over broken terrain is much harder than road driving and the follow behavior covers most of the actual requirement.

The load a dismounted soldier carries has grown to 40–60 kg, and that is a measurable cause of injury and reduced effectiveness. Moving even half of it onto a machine is a real improvement, and the same vehicle carries casualties and generates power for radios and drones. The weaknesses are terrain and noise. Slopes, mud, dense vegetation, and obstacles that a person steps over will stop a small vehicle, so the squad ends up waiting for or recovering it. Acoustic signature matters for a unit trying to move quietly.

Use logistics UGVs where the unit moves on foot over terrain a small vehicle can manage and the load burden is the limiting factor: dismounted infantry operations, mountain and arctic units, and resupply along routes too dangerous for crewed vehicles. If the terrain is genuinely broken, the vehicle will become a liability, and a legged platform or aerial resupply may work better. Judge these systems on how often the squad has to stop for the robot, which is the metric that has decided most field trials.

Payload 200–500 kg · speed 5–15 km/h · range 60–100 km per charge or tank · typical dismounted soldier load 40–60 kg · unit cost $150k–500k.

Small arms and fragmentation disable them easily since they are unarmored. Terrain does most of the work without any enemy involvement. Following a squad makes their route predictable. Jamming affects control and GNSS, though leader-follow using cameras is more robust than waypoint navigation. Their acoustic and thermal signature can reveal a unit that would otherwise be undetected.

General Dynamics MUTT and the US Army's S-MET program vehicles, Milrem THeMIS, Rheinmetall Mission Master, the Estonian and Ukrainian ground robots used for resupply and casualty evacuation under fire, and a growing number of Chinese designs.

Unit costs of $150k–500k are high relative to the load carried, and the business case has been debated in every army that has trialled them. Ukraine has provided the strongest operational evidence, using ground robots extensively for resupply and casualty evacuation along routes where drones make crewed movement lethal. That is a specific and demanding use case, and it has done more to justify the category than a decade of peacetime trials.

3 minMUTT: Rugged, All-Terrain MobilityGeneral Dynamics Land Systems · 10k+ views

2 minMilrem Robotics' THeMIS off-road autonomy demoMilrem Robotics · 100k+ views

4 minS-MET | An 8-wheeled Robotic Mule For Dismounted SoldiersUnited States Defense Media Channel · 1k+ views

The Army's Robotic Combat Vehicle (RCV) Program (Congressional Research Service)

A robotic combat vehicle is an uncrewed armored vehicle of 5–30 tonnes carrying a remote weapon station, sensors, or electronic warfare payloads, operating ahead of crewed formations. The concept is to put the first vehicle into contact without a crew in it, using the uncrewed vehicle to find the enemy, absorb the first engagement, and cue crewed vehicles that stay back. Control is by radio from a nearby crewed vehicle, with enough onboard autonomy to follow routes and hold formation when the link degrades.

Removing the crew from the most exposed position in a formation is the argument, and it is a good one — reconnaissance in contact is historically among the most dangerous tasks in ground combat. Uncrewed vehicles can also be designed without crew protection, saving weight and cost. The weaknesses are severe and have slowed every program. Off-road autonomy remains hard, control links in terrain are unreliable, weapons employment from an uncrewed vehicle raises unresolved policy questions, and the vehicles have cost far more than the concept assumed.

The case is strongest for reconnaissance and screening ahead of a crewed formation, and for carrying electronic warfare payloads that would otherwise attract fire onto a crewed vehicle. If the vehicle costs as much as a crewed one, the concept loses its rationale, so unit cost is the requirement to hold. For most armies the more affordable version of this idea has turned out to be small drones for reconnaissance and cheap ground robots for logistics, rather than armed armored vehicles.

Vehicle mass 5–30 tonnes · road speed roughly 40–80 km/h · endurance about a day of operations · control range from the crewed vehicle typically 1–2 km · unit cost in the millions per vehicle, comparable to a crewed vehicle.

Anti-tank weapons destroy them as readily as crewed vehicles. Control links in broken terrain are unreliable even without jamming, and jamming makes it worse. Terrain limits autonomous mobility. Because the vehicle is expensive and unarmored relative to a tank, the cost exchange against a cheap anti-tank weapon or an FPV drone is unfavorable, which is the same problem the crewed vehicles have but without the crew's judgment to mitigate it.

The US Army's RCV program with Textron and McQ, Milrem Type-X, Rheinmetall Mission Master SP in armed configurations, the Russian Uran-9 which performed poorly in Syria, and various Chinese and Israeli armed ground robot programs.

Programs in this class have repeatedly been restructured or canceled, most recently the US Army's RCV program, which was curtailed in favor of cheaper approaches. The pattern matches the stealth UCAV story: the uncrewed version ends up costing about as much as the crewed one, which removes the economic argument while leaving the technical and policy problems in place. The cheaper alternatives — small drones and light logistics robots — have delivered more operational value per dollar so far.

9 minThe Superiority of Milrem's Type X Robotic Combat VehicleMilitary TV · 10k+ views

2 minM-RCV - Medium Robotic Combat Vehicle with 30 mm turret - BL || Elbit Systems and DDR&D - IMODIsrael Defense · 10k+ views

4 minRussian Uran-9 unmanned combat vehicle tested in SyriaBinkov's Battlegrounds · 50k+ views

The Army's Robotic Combat Vehicle (RCV) Program (Congressional Research Service)

A legged ground robot walks on four legs, negotiating stairs, rubble, and broken ground that stops wheels and tracks. Commercial quadrupeds weigh 30–70 kg, carry 10–15 kg of payload, and run for 60–150 minutes per battery. Onboard autonomy handles balance, footstep planning, and obstacle avoidance, so the operator gives destinations rather than steering. Military and security use has focused on inspection and reconnaissance in structures, on hazardous-site survey, and on perimeter patrol, where the ability to climb stairs and step over debris is what makes the robot useful.

Mobility over terrain that defeats tracked robots is the unique capability, and it is genuinely useful inside damaged buildings, industrial sites, and tunnels. The autonomy stack is mature enough that operation needs little training. The weaknesses are endurance, payload, and cost. Two hours of walking is short, 10 kg is a modest payload, and $75k–150k is expensive for what is essentially a mobile camera. Legs are also mechanically complex and wear faster than tracks, and the robots are loud enough to be noticed.

Use legged robots where the terrain genuinely requires legs: stairs, rubble, industrial plant with catwalks and obstacles, and tunnels. If the ground is drivable, a tracked or wheeled robot does the job for a fraction of the cost with better endurance. If the task is aerial observation, a small drone is cheaper and faster. The honest assessment is that legged robots have found solid commercial footing in industrial inspection, where the routes are fixed and the value of avoiding a human visit is clear, and a narrower military niche than early enthusiasm suggested.

Mass 30–70 kg · payload 10–15 kg · endurance 60–150 minutes per battery · walking speed 1–3 m/s · unit cost $75k–150k for Western platforms, a few thousand dollars for Chinese quadrupeds.

Small arms disable them. Radio links fail underground and inside structures, which is where they are most wanted, so tethered or relay-based operation is common. Short endurance limits mission length. Their noise and gait make them conspicuous. Terrain that is genuinely hostile — deep mud, loose sand, ice — remains difficult for legs as well as for wheels.

Boston Dynamics Spot, Ghost Robotics Vision 60 used for base perimeter patrol by several air forces, ANYbotics ANYmal in industrial inspection, Unitree's much cheaper quadrupeds, and armed quadruped demonstrations that have attracted policy attention.

Chinese manufacturers have driven quadruped prices down by roughly an order of magnitude, with capable platforms available for a few thousand dollars against $75k+ for Western equivalents. That has expanded experimentation enormously while creating the same supply-chain concern as small drones. The durable commercial market is industrial inspection with a clear return on avoided human site visits; the military market remains smaller and focused on base security and hazardous reconnaissance.

8 minBoston Dynamics Spot Robot | All of its Engineering SECRETS!Sabin Civil Engineering · 5m+ views

20 minHow Boston Dynamics' Spot Robot Works!Adam Savage’s Tested · 1m+ views

3 minMILITARY TECH: Ghost Robotics VISION 60 Q-UGV Ground RobotMedia Magik Entertainment · 10k+ views

The Army's Robotic Combat Vehicle (RCV) Program (Congressional Research Service) · Research programs (DARPA)