Analysts frame the trajectory of military drones in eras. The current period — accelerated by the war in Ukraine — is often called the Third Drone Age, distinguished from earlier eras by three things: continued proliferation to new state and non-state actors, expansion across operational domains (air, land, sea, subsurface), and the arrival of AI-enabled autonomy on a “transparent” digital battlefield (ETH Zurich CSS).

A key conceptual shift: in earlier eras, drones were valued mainly for replacing troops in dangerous or dull roles in conflicts where one side had air superiority. In Ukraine, by contrast, “drones have boots” — they operate down to the lowest tactical level alongside infantry rather than substituting for them.

Within Ukraine specifically, the Hudson Institute identifies three phases:

• 2022 — Mass deployment. Ukraine’s crowdfunded Army of Drones pushed UAVs down to company level and trained thousands of operators; the drone became a tactical survival tool for reconnaissance and artillery adjustment.
• 2022–2023 — Strikes and counterstrikes. Both sides hardened air and electromagnetic defenses, causing massive fleet attrition. Medium-altitude drones largely disappeared from the tactical fight; loitering munitions and kamikaze systems came to dominate.
• 2023 onward — The FPV standard. First-person-view drones became the new norm, produced by the tens of thousands monthly and integrated into dedicated UAV companies within assault brigades. The battlefield became transparent to a depth of roughly 10–20 km.

By 2025, multiple sources estimate drones account for the large majority of battlefield casualties in Ukraine — a figure cited as high as ~80% (Hudson Institute, citing Army Technology).

2. Classification & Taxonomy

There is no single universal taxonomy; two systems are worth knowing.

Western / NATO-style classes (by weight and altitude):

• Class III — HALE (high-altitude long-endurance) and MALE (medium-altitude long-endurance) / combat platforms (e.g., Bayraktar TB2, Forpost, Orion).
• Class II — Tactical, sub-600 kg (e.g., catapult-launched fixed-wing recon/strike).
• Class I — Small, Mini, Micro (down to DJI Mavic/Mini-class quadcopters, FPV drones, Switchblade-class munitions).

Ukrainian functional classification. Rather than weight/altitude, Ukraine groups drones by structural configuration (quadcopter vs. fixed-wing), payload capacity, and — importantly — propeller diameter. Early-war FPVs were commonly 7-inch; demand for heavier payloads and longer range shifted the standard toward 9–10-inch quadcopters (CSIS/Bondar).

Functional types referenced across the documents:

• ISR — intelligence, surveillance, reconnaissance; the workhorse role enabling targeting.
• FPV — first-person-view rotary platforms derived from racing drones; fast, maneuverable, attritable.
• Loitering munitions / one-way attack (OWA) — e.g., ZALA Lancet (and cheaper Scalpel), Shahed-136.
• Bomber configurations — multirotors that release munitions from above (“Baba Yaga”-class heavy bombers).
• Relay drones — airborne signal relays extending range/control.
• Interceptor drones — low-cost air-defense drones used to down enemy ISR and strike drones (range ~30 km; cost ~$800–7,000; e.g., Sting, Shulika).
• “Middle Strike” — a category that emerged in early 2025; fixed-wing loitering munitions striking at depths of ~30–200 km.
• Deep Strike — long-range OWA drones reaching 1,000–2,000+ km, used against airfields, depots, oil/energy infrastructure, and defense industry (e.g., UJ-26 Beaver, AN-196 Liutiy, FP-1).
• Drone-missiles — jet-powered deep-strike systems reaching ~900 km/h, blending UAV scalability with missile kinetic energy.

3. UAV Technology & Development

Platform architecture

At a component level, the modular small-UAV stack typically comprises a flight controller (FC), electronic speed controllers and brushless motors, a power system (lithium-polymer batteries or, increasingly, hybrid/alternative power units explored in the research literature), a video transmission system and FPV camera, and the control/communications link. The modularity of this architecture is central: a single base quadcopter can be reconfigured for ISR, relay, bomber, or kamikaze roles by swapping payloads and components (CSIS). That same modularity, RUSI cautions, creates a misconception that mass production is simple — the real challenge is replicating quality at high volume and low cost.

Modular software & “componentized” autonomy

A defining 2023–2025 trend (CSIS/Bondar) is decoupling capability from any single airframe: Ukrainian firms build standalone modules — a chip with embedded software, sometimes a camera, often smaller than a bar of soap — that add functions such as target recognition or “last-mile” terminal navigation, and can be integrated across FPV drones, ground vehicles, and even turret-mounted weapons.

Autonomy & AI

Three functions are advancing toward greater autonomy: ISR, automatic target recognition (ATR), and autonomous navigation. By 2022 basic automation (takeoff/landing, course/altitude hold, waypoint routes) was taken for granted. By 2023–2025, some systems could visually acquire and track a target and guide themselves toward it, with effective acquisition ranges typically 500–5,000 m depending on the electro-optical system and algorithm robustness. Sensor/compute modules are now capable enough to support target detection, recognition, coordinate transmission, and handover between drones — the building blocks of swarming, already in testing in Ukraine.

The company The Fourth Law describes a five-level autonomy roadmap: (1) automatic terminal “last-mile” guidance, (2) automatic target engagement, (3) autonomous target selection, (4) autonomous navigation, and (5) full autonomy. Other notable players include Swarmer (swarm coordination), Bavovna.ai (a hybrid inertial navigation kit fusing accelerometer, gyroscope, compass, barometer, airflow sensor, and AI to provide positioning independent of GNSS and resistant to EW), and Twist Robotics (visual navigation, plus the Obriy FPV simulator).

Electronic-warfare resistance

Because the current generation of drones is heavily contested by jamming, two engineering responses recur: frequency agility (changing frequency in flight) and machine-vision terminal guidance (so the drone needs no operator link on final approach). The most prominent hardware response is the fiber-optic-controlled drone, which trades range/weight for near-immunity to RF jamming by using a physical fiber tether for control and video — a major 2024–2025 battlefield development (multiple sources). One CSIS observation captures the strategic tension: EW defeats today’s drones, but won’t work once full autonomy arrives — “we must invest in full autonomy if we are to counter the counter.”

4. The Drone Industrial Ecosystem

Two contrasting production models dominate the documents.

Ukraine — decentralized, COTS-driven mass

Ukraine’s model is built on commercial off-the-shelf components, rapid iteration, and a sprawling base of producers. Government coordination runs through Brave1 (which signals frontline needs, awards grants, and coordinates priorities) and the Army of Drones program. Production scaled from tens of thousands of units per month in 2022 to hundreds of thousands by 2024; Ukraine manufactured well over 1.5 million FPV drones in 2024 and roughly 2 million drones overall, with a stated ambition far higher. One analysis counts 500+ identified manufacturers, from individual workshops to commercial enterprises, and per-unit FPV build times of just 2–8 hours. The innovation cycle itself is extraordinarily compressed — the Australian Army paper puts it at one week to ~three months between new solutions or significant modifications.

Russia / Iran — centralized, state-directed

The Shahed/Geran line represents the opposite philosophy: centralized state-directed manufacturing (notably the Alabuga facility in Tatarstan, established via a ~$2 billion 2023 Iran–Russia deal involving technology transfer) scaling through industrial policy rather than commercial markets. Russia simultaneously caught up in cheap racing-drone strikes after an initial period of underinvestment.

Production economics — “cost-per-effect”

The economic logic is the recurring theme:

• FPV drones: ~$300–500 per unit (mean ~$400), built from consumer-grade components.
• Loitering munitions (Lancet-class): ~$35,000–50,000.
• Shahed-136 / Geran-2: ~$20,000–50,000 (Iranian estimate) to ~$48,000–80,000 (Russian-built variant) — versus ~$1.5 million for a comparable-range cruise missile.

This inverts conventional air-campaign logic: when the platform is attritable and replaceable, its individual cost matters far less than its one-way effect, and cheap mass can overwhelm defenses so that fewer expensive munitions get through. RUSI and the Australian Army both stress the counter-side of this equation — defenders must “flip the economics” so that destroying a drone costs less than the drone itself.

5. Supply Chains & Components

This is one of the most strategically important threads in the corpus (RUSI’s Decoupling Supply Chains from China is the anchor source).

• China’s position: China supplies roughly 80% of the global small-UAS market (complete systems and components) and produces around 90% of global neodymium, the rare-earth element behind the permanent magnets in brushless motors.
• Pressure points identified: brushless motors and magnets (REE dependence); flight controllers (China’s SpeedyBee long dominated use in Ukraine; processors from STMicroelectronics and competitors like ArteryTek); video-transmission chips (Taiwanese-purchased chips can cost up to ~10× DJI-designed equivalents); cameras and sensors (Sony CMOS sensors, plus germanium — largely China-sourced — for infrared/thermal lenses); and gimbals (historically DJI-dominated; new entrants like Gremsy in Vietnam).
• Cost of decoupling: “non-Chinese” builds can run “ten times the cost” for magnets, and NDAA-compliant FCs were priced over ~$90 (plus a comparable ESC) versus ~$70 SpeedyBee stacks in 2025. A 2024 U.S. Chamber of Commerce survey found 44% of surveyed circuit-board manufacturers couldn’t determine whether their products contained PRC-foundry chips.
• Decoupling efforts: the U.S. Defense Innovation Unit’s Blue List, Australia’s Sovereign UAS Challenge, Taiwan’s Drone National Team, and partnerships such as Auterion–Ukraine (the “Skynode”/AI strike-kit co-production). RUSI’s three identified opportunities: scale capacity through coordinated allied production, deepen international innovation cooperation, and find a middle ground between cheap COTS mass and standardized higher-cost systems — while warning that resilient supply chains take years to build.
• The “Shenzhen problem”: the U.S. and allies lack the industrial clustering (suppliers, engineers, logistics co-located) that gives China’s drone hub its network effects and cost/innovation advantages (KSE Institute).

6. Operational Employment & Battlefield Effects (conceptual)

• The transparent battlefield. Ubiquitous ISR has made concealed force build-up — the precondition for breakthroughs against prepared defenses — extremely difficult, contributing to positional, attritional warfare (IISS). Analysts draw comparisons to the firepower-favors-defense stalemate of early WWI.
• The reconnaissance-strike complex. Drones are the critical link in the network that acquires, processes, and transmits targeting data to fires. Paired with artillery, they compress targeting timelines dramatically — detection-to-strike reduced to minutes in some cases (Hudson Institute; CSS).
• From kill chains to kill webs. Rigid linear targeting is giving way to decentralized, data-driven “kill web” architectures resilient to attrition, often using stacks of drones with distinct functions sharing the same airspace under agile command and control.
• Adaptation and counter-adaptation. As ISR/strike saturation challenged massed mechanized assault, Russia shifted toward dispersed small-unit and motorcycle-mounted actions and “turtle tank” improvised armor — none of which has restored massed tactical mobility (IISS).

7. Counter-UAS (C-UAS)

The unifying doctrine is layered defense — multiple engagement opportunities beginning at maximum range, before an attacking UAS can release its weapons (U.S. Army doctrine, ATP 3-01.81). “No single bullet” solves the problem.

Detection layer

A robust detector fuses complementary sensors (IEEE Nation’s Defense review): radar (long range, but can confuse small drones with birds), RF sensors (detect the drone–operator link; effective against COTS drones), electro-optical/infrared (EO/IR) cameras (visual confirmation, low-visibility tracking), acoustic sensors (propeller signatures; short range, useful in urban settings), and AI-based software that integrates all of the above to improve accuracy and filter false positives.

Neutralization layer

Approaches span:

• Electronic countermeasures / jamming — disrupt the control and navigation links (the current mainstay; increasingly defeated by fiber-optic and autonomous drones).
• Spoofing — feed false navigation/control data (e.g., D-Fend EnforceAir, Skyfend Spoofer).
• Kinetic effectors — guns, nets (including net-launchers like SkyWall), and interceptor drones.
• Directed energy — high-power microwave (HPM) systems such as AFRL’s THOR and Epirus Leonidas (effective against swarms), and high-energy lasers such as the U.S. Navy’s HELIOS, Rafael’s Iron Beam, and the British Army’s vehicle-mounted HEL trials (2024).

Organization & force structure (U.S. example)

The Joint Counter-small-UAS Office (JCO), established 2020, endorsed eight interim systems and continually re-evaluates products. Air Defense Artillery and Low Altitude Air Defense personnel operate platforms like LIDS (Army) and MADIS (Marine Corps). Under the 2024 “Force Structure Transformation,” the Army planned additional SHORAD and IFPC battalions and nine dedicated C-UAS batteries — though no service has yet created a dedicated C-UAS military occupational specialty. The FY2025 NDAA directed a large-scale C-UAS exercise.

The core difficulty: cost-exchange

Defenders cannot sustainably trade million-dollar interceptors against thousand-dollar drones. The strategic imperative is cheaper effectors and saturation-oriented doctrine — the documents repeatedly frame “winning the cost-exchange” as the central C-UAS challenge. The IEEE review also flags a military-vs-civilian divide: military systems enjoy latitude on jamming/spoofing and kinetic kill, while civilian deployments are constrained by spectrum law, airspace regulation, privacy, and liability.

8. Military Concepts & Strategic Implications

• Revolution or not? Several authors argue the lesson mirrors the tank after WWI: the revolution lies not in the platform but in the doctrines, organizations, and industrial strategies that integrate it. The risk is fixating on hardware while missing the operational-art shift toward network-centric, multi-domain warfare.
• The doctrine gap. Traditional militaries lag in doctrine for mass/attritable systems — how they integrate with legacy platforms, how attrition is managed, how rapid reconstitution works, and how autonomous systems are commanded and controlled.
• Deterrence and posture. Analysts argue deterrence must be re-examined in light of cost-exchange dynamics: if adversaries believe they can impose unsustainable costs through mass drone use, deterrence can fail. Alliance burden-sharing metrics may need to credit contribution to aggregate attritable capability, not just GDP-percentage spending.
• Procurement reform. The recurring prescription (CSS, Australian Army, Hudson): faster acquisition matched to a weeks-to-months innovation cycle; outreach to non-traditional/private-sector suppliers under flexible contracts; and unit cost/quantity appropriate to a low “cost-per-effect” world. Sovereign capability, an R&D hub for UAS/C-UAS, and continuous learning from Ukraine are emphasized.
• Combined-arms caveat. Drones augment rather than replace existing systems. As the Australian Army paper stresses, infantry still need tanks, tanks and infantry still need artillery, and the capacity that often mattered most in 2022–2023 was artillery ammunition and infantry — not drones alone.
• Unresolved ethical/legal questions. Accountability for autonomous or accidental engagements; whether low-political-risk strike lowers the threshold for force; and the human-rights dimension where drones have been used against civilians.

9. Key Lessons & Outlook

1. Iterate at the speed of the threat. Edge is held by week-to-month adaptation cycles, not multi-year programs.
2. Mass and attrition are features, not bugs. Cheap, replaceable, modular systems generated through componentized software and COTS hardware define the model — and the counter-model must be equally economical.
3. Autonomy is the next escalation. As EW matures, machine-vision terminal guidance, inertial/visual navigation, and swarming move capability past the jamming era; full autonomy is the stated destination, carrying the heaviest ethical weight.
4. Supply chains are the strategic chokepoint. Whoever controls magnets, mature chips, sensors, and the clustering to mass-produce them at quality holds leverage; decoupling from China is achievable but slow and costly.
5. C-UAS is a layered, multi-domain, never-finished problem. Detect-decide-defeat across RF, optical, acoustic, kinetic, and directed-energy layers — optimized for cost-exchange.
6. The revolution is doctrinal. Integration into multi-domain “kill webs,” not the airframe itself, determines whether tactical disruption becomes lasting strategic effect.

Appendix: Source Documents

The compilation draws on the project library, including:

• Drones in Modern Warfare: Lessons Learnt from the War in Ukraine — Australian Army Occasional Paper No. 29
• Learning from the Ukrainian Battlefield — ETH Zurich CSS Study (2024)
• From the Battlefield to the Future of Warfare: Harnessing Ukraine’s Drone Innovations — KSE Institute (Nov 2025)
• Ukraine’s Future Vision and Current Capabilities for Waging AI-Enabled Autonomous Warfare — Kateryna Bondar / CSIS (Mar 2025)
• Drones: Decoupling Supply Chains from China — Robert Tollast / RUSI
• Department of Defense Counter-UAS: Background and Issues for Congress — Congressional Research Service (R48477)
• UAVs: ISR, Deterrence and War — IISS Strategic Dossier (Ch. 3)
• The Impact of Drones on the Battlefield — Tsiporah Fried / Hudson Institute
• Nation’s Defense: A Comprehensive Review of Anti-Drone Systems and Strategies — IEEE Access
• Analysis of Low-Cost Drone Warfare (Ukraine/Iranian Shahed programs, 2022–2026)
• Ukraine Drone Ecosystem Analysis (2025); plus additional technical, tactical, and review papers in the library.

This document is an analytical/educational synthesis intended as website reference content. Figures and unit costs are drawn from the cited sources and reflect 2024–2025 estimates; verify against primary sources before publication.