Features of Air Power
The side in control of the air tends to win. At a minimum, dominant air power is a massive force multiplier that allows the side wielding it to take significantly less casualties than its opponent. Aircraft can uniquely disrupt supply lines, command and control, and troop concentrations. The forces on the losing side must drastically alter their tactics to survive, limiting their ability to attack or defend.
Another feature is that air-to-air battles tend to be lopsided. It is more common to see 20:1 or 10:1 kill/loss ratios than even matches. For example, the F-15 has 104 kills and zero losses since entering service in 1976. The defining factors have been pilot quality, aircraft performance, weapon performance, and sensor capability (radar, airborne early warning aircraft, etc.).
US airpower was so dominant in the 20th century that most opponents focused on building ground-based anti-aircraft defenses. An arms race developed between these anti-aircraft missile batteries and ever more sophisticated aircraft, weapons, and tactics on the US side. Stealth to avoid detection, cruise missiles to avoid risking aircraft, and highly specialized tactics and weapons to defeat anti-aircraft batteries are an outgrowth of this competition.
Drones add a layer of complexity and prompt questions about how they fit into the traditional airpower paradigm.
What Drones Change
The most accurate paradigm for today's drones is that they are not a new category but dramatically reduce the cost of some existing functions. For example:
-
FPV Drones -> Attack Helicopters
Advocates of rotor aircraft thought they would dominate the battlefield in the 60s, 70s, and 80s, to the detriment of traditional armor. It didn't happen because helicopters are vulnerable to air defenses, including shoulder-fired missiles and anti-aircraft guns.
First Person View (FPV) kamikaze drones that cost <$1000 or slightly larger reusable drones are bringing this prediction back from the dead. They are still vulnerable to air defense, but it is irrelevant given their cost. Ground forces will need to make many adjustments, similar to when anti-tank guided missiles made WWII-style tanks obsolete in the 1960s and 1970s.
-
"Bomber Drones" -> Attack Helicopters pt. 2
Some missions call for slightly larger munitions than disposable FPVs can justify, and "bomber" drones that weigh around 25-50 kg and cost $10,000 fill the void. They mostly fly at night to increase survival rates and often use satellite communications, like StarLink, to avoid jamming. Missions are attacking parked vehicles, mining roads, and dropping grenades on infantry. These drones are much more powerful than FPVs and are worth the price if they can survive a few missions.
-
Recon Drones -> Scout Helicopters and Forward Air Control Aircraft
Scouting for artillery, ground attack aircraft, and attack helicopters has long been a scarce resource, even for the US military. Infantry and armor units still had to self-scout with limited visibility.
Small recon drones, often off-the-shelf commercial models, bring top-tier scouting down to the squad level. Their cost makes using them sustainable, while many large drones, like the US Predator, are obsolete in high-intensity battles because of their price and vulnerability to air defenses.
-
One-Way Attack Drones -> Cruise Missiles
Cruise missiles have a unique ability to attack heavily defended targets in the opponent's rear, but their price limits their number.
Propeller-powered one-way attack drones can cost as little as $50,000 instead of $1+ million, increasing volume. The overall impact has been much more muted than FPV and recon drones because these drones are so easy to shoot down and have small payloads that limit what targets they can be effective against. They travel slowly, roughly the same as a car on the interstate, to meet cost goals and extend range. Their utility plummets once the opponent adapts to shoot them down with cheap weapons, like guns on trucks and helicopters, cheap interceptor drones, or electronic warfare. The drones can still provide net benefits if they temporarily overwhelm air defenses, force the enemy to expend significant organizational resources to counter them, or the targets are valuable enough.
-
Interceptor Drones -> Man Portable Anti-Aircraft Missiles
Militaries developed man-portable anti-aircraft missiles to counter helicopters and low-flying aircraft, but they are much too expensive and complex to counter drones.
Instead, small racing-style drones that cost no more than a few thousand dollars ram or explode near targets. Their prey is primarily more expensive attack drones and higher-tier recon drones that cost $30,000-$200,000.
There are some experiments with drones carrying shotguns and other air-to-air weaponry to deal with the smallest FPV and recon drones. Time will tell if these are viable.
Some categories haven't seen drones encroach on their territory, namely air-superiority fighters or bombers.
The main battlefield effects are that low-level rotor-based attack is effective, detection of movement is much more likely, integrated air defense systems have to handle an order of magnitude more bogies, and slow speed, low altitude aircraft are becoming vulnerable to small interceptors.
Analyzing Trends
The rapid development of these categories tells us several things about the future of drones:
-
Barbell Procurement Strategy
The battlefield is so hostile that drones must be cheap enough to be expendable or capable enough to avoid all air defenses. The somewhat fancy $100,000 recon drone is probably in no-man's-land. Large drones without sophisticated countermeasures, like the US Global Hawk or Predator/Reaper family, are obsolete outside the most permissive airspace. Even drones that were considered cheap before the war in Ukraine, like the Turkish TB-2, have been sent to the scrap heap.
One of the only viable(?) large drones currently in use is the pricey US RQ-180 because of its size and modern stealth features. Traditional cruise missiles also continue to be viable for deep strikes.
-
Small Eats Large
Drones aren't automatically cheaper than legacy systems like helicopters or strategic reconnaissance platforms. Radical reduction in size and complexity is the best way to achieve this.
Better electronics and cameras have allowed recon drones with mass measured in grams. Or a shaped charge driven by an FPV drone into the weakest part of a vehicle's armor can be much smaller than traditional anti-tank missile warheads.
Battery-electric powertrains can shrink much more than engines can, and these drones have disrupted short-range, low-speed categories much more than long-range or high-powered missions.
The success rate of these drones is often low, between 10%-50%, and many targets need multiple hits. However, the low cost of small drones means the math is favorable and similar to artillery shells.
-
Single Function Dominates
Drones can only be small and cheap if they are highly specialized for one task. Examples include anti-vehicle, anti-personnel, high-value targets ~30 km behind enemy lines, hitting enemy drones, dropping mines or supplies, etc. Many of these categories even have further specialization within them.
-
Paths with Faster Iteration Win
Things change fast since small drones are a relatively new technology. Pathways that allow quick adjustments can outcompete slow paths. Small and single-function platforms can increase iteration speed.
The Drone Air Force
Drones have changed the battlefield by supercharging army rotor aviation functions and organic air defense. They only partially substitute for traditional air power functions like close air support or air defense/superiority.
A true drone air force must keep the lessons of the barbell, small, single function, fast iteration, and proper powertrain in mind to revolutionize air power. Otherwise, it will be a traditional aircraft without a pilot. The designs must provide cheap precision, be reliable, and be easy to mass-produce to stay relevant in high-intensity conflict.
The single function requirement provides the guide for what models are needed. The designs assume efficient and tightly integrated designs - high payload/battery fraction, high lift-over-drag ratio, etc. The SpaceX/Tesla treatment rather than the catalog engineering special.
Airborne Early Warning (AEW)
Radar signal strength diminishes rapidly, requiring powerful radars to see long distances. Those radars are too bulky and expensive to put on each fighter. Air forces convert civilian airliners to radar aircraft instead. Opponents can usually detect radar signals at double their range, so it's helpful for fighters to keep their radars off to conceal their positions and get directions from the airborne early warning (AEW) aircraft.
These legacy AEW platforms are slow, expensive, and obsolete in high-intensity warfare. Ukraine has knocked out several of Russia's without an effective air force. Long-range missiles with final radar guidance are the kiss of death for planes like the US E-2 or E-3.
The US Air Force wants to distribute this capability in $10-$30 million drones that fly as "loyal wingmen" to manned stealth fighters. The procurement cost for the hundreds of these drones needed to fulfill the same function as 24/7 coverage from traditional AEW platforms would be similar to buying E-3 replacements.
But the drone success algorithm might produce something better. Passive sensors are lighter, cheaper, and use less power, but they also have short detection ranges and are more sensitive to weather. It might require 5000-10,000 drones to replace one 24/7 circuit. However, it is an application that can work for electric batteries, meaning the total cost will probably be 1/100th of jet drones or traditional AEW platforms.
The drones would be fixed-wing, powered gliders weighing a few kilograms, optimized to circle a small area for 6-10 hours. They would hunt for enemy aircraft or missiles using cameras, infrared, and microphones and send bursts of information if their onboard processing notices anything interesting. The cost would be <$5000/unit to absorb losses from interceptor drones. Their sensors, batteries, and software will improve much faster than the multi-billion dollar development projects that are the alternative.
Air/Missile Defense
Air defense today usually means stopping incoming cruise missiles and drones rather than dogfighting. Tactical fighters like the F-16 are capable but overkill for the role. Aircraft are valuable in this role because their mobility allows them to deploy more places and get close to their targets. Close engagements require less powerful radars and smaller, cheaper missiles than a ground-based battery covering a large area.
The ideal munition to base the vehicle around is probably the APKWS rocket or another comparable short-range missile/rocket like the Zuni. It is mass-produced, cheap, and accurate as long as it's within a few kilometers of the bandit. It is proven against larger drones in Ukraine and successfully tested against cruise missile-type targets.
An off-the-shelf option is putting missiles or rockets on the Kratos XQ-58, which costs $5-$10 million kitted out. Its range and payload are overkill when opponents plan missile and drone salvos to have their weapons converge from multiple angles, limiting how many one aircraft can intercept.
The "safe" answer for propulsion is jet engines. High subsonic speed is the minimum to counter cruise missiles with a decent interception rate. Liquid fuel enhances the loiter time and increases the area the aircraft can defend. Anduril's Roadrunner is an example of what the aircraft might look like. They sell them for ~$500,000 each, including some auxiliary systems.
A battery-electric drone architecture would be an exercise in brute force. The fastest electric drone is around 300 mph and not quick enough for conventional intercepts. Aircraft power increases at the cube of velocity, so faster drones can end up on the wrong side of the cost curve. Instead, the strategy would be multiple "picket" lines of slow, tightly spaced drones. The APKWS only weighs 15 kg, and available radar models with enough range to guide small rockets/missiles weigh <20 kg . The airframe and propulsion can be flimsy and cheap because of the small payload and slow speed. The main challenge would be making the cost of the avionics package reasonable. These would exclusively operate over friendly territory to maintain survivability.
The algorithm would prefer the brute force, slow, and cheap drone strategy. Over time, better batteries and motors should improve performance while avionics costs fall. The logistics tail and basing requirements would be minimal compared to jets, and the drone would ideally be cheap enough to keep adding picket lines as necessary. It is a rapidly scalable air defense that is more cost-effective than fixed missile batteries and more challenging for opponents to game plan due to the inherent mobility.
Air Superiority
The current thinking is that modern air-to-air engagements between top-tier air forces begin like a Wild West duel. Each side rips off salvos of long-range missiles, hoping to disable their opponents first. Engagement distances might remain long with no visual range combat. One side in any hypothetical engagement almost certainly has an advantage in front-line stealth fighters, pilots, and missiles that would end the air battle relatively quickly.
The requirements for an air superiority fighter are locating opponent aircraft and firing weapons that can get a hit. The point of fighters like the F-22 is to make this difficult. They are hard to detect, challenging to get a missile lock on, have exceptional agility, and can push over Mach 2. The US air-to-air missiles, which presumably have a chance at a target like this, range from $400,000 to several million dollars. And it takes a skilled pilot in a capable aircraft to get in position for a high-quality shot.
The US Air Force's air superiority drone strategy is relatively capable models ($10-$30 million) that carry more missiles and scout for manned fighters.
A risk with our cheap drone algorithm is that stealth fighters can load up with dozens of cheap air-to-air missiles that easily handle slow, visible drones. The fighters can rip apart drone swarms without any drone being able to get a firing solution because the drones only have short-range radars, and swarm intelligence often provides qualitative instead of quantitative vectors. The cost could be upside down since these won't be $5000 drones, plus each one carries a missile costing >$400,000.
A cheap, battery-electric air superiority drone has to cheat. A small, slow drone is easier to make stealthy than a supersonic stealth fighter. There are typically tradeoffs between stealth and maneuverability, like not having a tail, but losing maneuverability is fine here. The small size helps further reduce the radar cross-section and visibility. The thermal signature would be minimal with such an efficient powertrain. And radar-absorbing paint doesn't have to last long or be the best. Decoys and clutter-like chaff will be much more effective when the drones are slow-moving. A drone like this could also have a very damage-resistant design compared to a fighter that is mostly fuel and a massive engine spinning at thousands of RPM. The low signature would hinder the opponent's ability to locate and kill the drones with simple, cheap missiles, while durability increases the opponent's required missile size/amount.
Flying at very slow speeds would also provide the range and loitering time to patrol forward of friendly lines, though it would not be able to operate from rear bases. Engagements would be extremely close range and facilitated by the sheer number of drones spaced every few kilometers. Visible light, IR, and synthetic aperture radar sensors would have a chance of detecting opponent stealth fighters at short range. Air-to-air missiles can be less sophisticated because they need minimal onboard guidance for close-in engagements, and the target aircraft would only have a few seconds to maneuver.
The strategy would be more like "air denial" than air superiority. American-style air dominance is aggressive, chasing down bogies while accompanying aircraft attack air bases. A slow-moving cloud of drones is easy to avoid but creates a no-go zone.
There is a good chance the low-end drone strategy doesn't work in air-to-air combat, given the history and how extreme the evolutionary pressure will be to locate and out-range these no-man's-land cost drones with cheap munitions. Achieving one kill for 100 or 1000 losses is not given against modern fighters, nor is one kill for 10 to 100 losses against low-end jet drones. The evolutionary successful strategy could be an AI pilot flying a $200 million fighter. Cheap air denial drones could still be helpful in creating temporary bubbles for other aircraft to work. It is possible that traditional air superiority aircraft lose because they are vulnerable on the ground and dependent on support aircraft (tankers, airborne early warning) rather than any deficiencies in the air.
Suppression and Destruction of Air Defense
The early days of neutralizing surface-to-air missile batteries were harrowing work. Attacking aircraft would fly right above the treetops and pop up to fire a radiation-seeking missile or drop cluster munitions. The sites were simple to disable because missiles needed central guidance to the targeted aircraft. Radars were single-point failures. The evolution has been toward more capable surface-to-air missiles and independent subunits, making them much more resilient. Attackers now use longer-range, fast anti-radiation missiles with multi-mode seekers to track down vehicles that cut and run after seeing a missile launch. These methods still seem adequate and effective for high-end, non-Western air defenses.
Towed anti-aircraft artillery is one obvious counter to slow drones. These guns are cheap, easy to repair, don't require precise target location, and can handle high volumes of drones. Drone formations would need a new class of cheap missile/loitering munition for these weapons. They probably look like Anduril's Altius series of drones, but 10x cheaper.
Tactical Bombing
Glide bombs have become a staple of US aviation and one of the most effective weapons in the Ukrainian war. They are cheap, accurate, and provide enough stand-off range to increase the survivability of the launching aircraft. Their size (250-2000 lb) packs more punch than small drones or tube artillery. Air superiority allows tactical bombers wielding the bombs to wreak havoc behind enemy lines. They can pound enemy positions and forward command posts near the line of contact if enemy airspace is well-defended.
The off-the-shelf option for a drone tactical bomber is, again, the Kratos XQ-58, which can carry four small diameter bombs with a several thousand-kilometer range.
The algorithm would suggest picking a smallish glide bomb, basing the airframe design around carrying a single bomb, and checking if an electric powertrain can meet the range requirements.
The output looks something like an aircraft that weighs 300 kilograms and carries a 120-kilogram small diameter bomb. A battery-electric powertrain could give it 150-300 kilometers of combat radius, plus 20-100 km from the bomb's gliding. Flight speed would be relatively slow, but it would accelerate briefly at release to give the bomb more velocity or at takeoff to use short fields/roads. The battery would be about 1/5 the capacity of a Model 3, and the drone could fit in the bed of an F-150 if the wings fold in. The range is relatively conservative because such a drone could use battery cells with no reserve and minimal fire safety systems. The vehicle might barely cost more than the bomb it carries($50,000), so some losses are acceptable, but the general flight path is low risk because of the stand-off from the glide bomb.
That capability would be relevant in Ukraine, Syria, Taiwan, and other possible theaters. The development cost could be low, so it would not be a loss if the design became obsolete. And US allies could operate these aircraft. Moving tactical bombing to small, electric platforms that don't need airfields simplifies the logistical tail, reduces reliance on vulnerable and politically contentious regional bases, and provides organic tactical bombing without constant overhead coverage from traditional airframes.
Sea Mine Laying
Bomber-dropped mines were one of the crucial blows against the Japanese in WWII and could have similar impacts against some potential US opponents. The trouble is dropping them in contested airspace with enough volume to matter. One recent innovation has been to turn glide bombs into mines, providing stand-off range for aircraft.
Defensive mines in the shallowest water and harbors, like for Taiwan or the Philippines, are easy, along with compact regions like the Persian Gulf or Baltic Sea. The solution looks like the tactical bomber.
Slightly larger battery-electric drones can mine most of the Chinese coast north of Taiwan, flying from South Korea, Taiwan, or Japan.
Southern China is much harder to access and would likely be handled by sophisticated, multi-platform air operations or other means, like submarines.
Strategic Bomber
Mass and range matter with bombers, making a cheap drone paradigm challenging. Bombers are most important in contested airspace because tactical aircraft can handle most attacks once the "wild weasels" degrade air defenses. Performance is required. The US B-21 meets the need.
One logical alternative to stealth bombers is concepts like the US Air Force's Rapid Dragon, which allows cargo planes, like the C-130, to launch cruise missiles. The C-130 is still in production after decades, has a good payload and range, is cost-effective to buy and operate, and can take off from rough airfields. The plane is a drone carrier with cruise missiles being the drones.
A cutting-edge replacement would be reusable hypersonic drones. These aircraft would use speed to avoid enemy defenses. Munitions can be simpler and cheaper than cruise missiles because the glide distance of a bomb launched at Mach 5 from 70,000 feet can be hundreds of kilometers. The game-breaking speed that outruns missiles and increases standoff range allows a hypersonic drone to stay within the algorithm. Engineers can focus on making it fly fast instead of traditional fighters that require sophisticated engineering across many domains to remain alive for more than a few minutes in hostile airspace. An aircraft like this would be disruptive by allowing attacks in the enemy rear before gaining air superiority and potentially having a much higher sortie rate than slow stealth bombers.
Munition Innovation
Today's munitions are optimized for the pre-smartphone era, often having expensive electronics built to military specifications that aren't commercial-off-the-shelf friendly. Guidance kits for rockets and bombs turn $500-$3000 munitions into $30,000-$50,000 ones. That is a steal compared to million-dollar cruise missiles but ridiculous compared to what could be. Incorporating computer vision and other cheap sensors seems like low-hanging fruit to lower cost.
There also seems to be some convergence on cruise missiles/drones that are a tenth the size of traditional cruise missiles with similar speed and less range. These require new, small jet engines, and the diversity in effort is remarkable. Several groups are dusting off the German WWII V-1 pulse jet. Others like Kratos or Ukrainian groups are leveraging their years of making cheap engines for target drones or other lower-tier applications. And some, like Anduril, are developing clean-sheet designs that they hope to use in many applications.
New classes of bombs, missiles, and sea mines optimized for drones would amplify effectiveness.
Manufacturing and Logistics
The creation and operation of a drone air force differs from modern air forces.
Operations
The slow, electric drones have massive advantages in operating footprint and costs, with that impact reverberating down the logistics tail:
-
Fuel
Fuel is one of the biggest concerns for modern militaries. The US military often assumes a fuel cost of hundreds or thousands of dollars per gallon to deliver to war zones for planning. The volume of demand in a high-intensity conflict could reach the level of economies like Japan. Aircraft are often the largest fuel consumers.
These drones require a fraction of the energy of high-performance aircraft and can often use electricity instead of fuel. The AEW scout battery pack would only be a few pounds and could charge with a tiny solar panel. Models like the tactical bomber, air-defense fighter, or air superiority fighter have batteries small enough to swap by hand. A few standard-size solar panels could provide enough juice for one sortie per day and don't require vulnerable centralized infrastructure.
-
Infrastructure
An aircraft's weight and stall speed plays a large part in determining runway length and quality. Small, slow drones need minimal airstrips, if they need them at all. There is no need for traditional air bases.
-
Parts/Maintenance
The US military prefers "module-based" maintenance. Techs change an entire radar module instead of diagnosing and fixing a certain subcomponent to reduce labor hours and the number of parts in stock.
Many drones would cost as much as a typical module, and there would be no reason to bother with parts or repairs. The need for techs and parts management would be minimal.
Battery-electric powertrains are reliable compared to jet engines and should be able to fly hundreds or thousands of hours before replacement without maintenance.
-
Training
Fighter pilots are the most valuable rank-adjusted human capital in any military. One great pilot can make a meaningful difference in an entire war by helping to clear the skies. Selection is intense, and training is very slow. Simulators help, but learning to fly a $100+ million fighter jet doesn't happen overnight.
AI pilots take more effort to train initially but can replicate as needed. The burden for any human pilot/manager will be lower given the narrower mission of drones than multi-role fighters. The low cost of the platforms means both AI and humans can train constantly on real aircraft instead of using simulators. Battles can be live instead of simulated without endangering human pilots, improving the quality of training.
-
Sortie Rate
Most fighters need full crews to turn the aircraft around and keep it flying. Each airframe only has so many hours without full refits. Many aircraft struggle to fly one sortie per day. A low-maintenance, battery-electric drone with swappable batteries could fly 20-22 hours each day.
The sortie challenge would be especially beneficial for countries like Taiwan. China constantly flies fighters at the edge of Taiwan's air defense identification zone, which forces Taiwan to send fighters to intercept them, wearing down airframes and pilots. A constant picket of drones would negate this strategy.
-
Shipping
Munitions, especially bombs, are the biggest logistical challenge after fuel. Manned aircraft tend to drop large bombs that are overkill because of their limited sortie rates and the risk each mission entails. Drones with high sortie rates can use small bombs that make the drone more practical and reduce total tonnage dropped as each target gets the appropriate amount instead of a truck getting vaporized by a 2000 lb bomb.
Battery-electric drones are a dream from an operations and logistic standpoint. High sortie rate, no fuel deliveries, low cost per flight hour, no maintenance tail, and reduced munition tonnage are all features.
Manufacturing
Today's militaries build aircraft in small numbers (<1000), with production runs lasting a decade or two. Only the most popular models, like the F-16 or the C-130, stay in production and reach thousands of units. Airframes last decades and undergo upgrades - new radars, wings, etc. - to extend their life. The upfront expense for new aircraft is eye-watering, and the fixed cost is high because of the low production numbers.
A drone air force would be the opposite. The development cost for single-purpose drones with battery electric power trains would be relatively low. Production volume would be extremely high. And the lifetimes of the drones would be short with no upgrades.
Production needs to absorb lessons from other high-volume programs, like munitions. Demand skyrockets during wars, but production lags because factories take time to come online. The two main strategies to deal with this are stockpiles and underutilized factories.
Stockpiles are common for artillery shells where manufacturing equipment is expensive, designs rarely change, and the shells have long shelf lives. Drones change fast, and production is more akin to light manufacturing, meaning extra factory capacity is a better strategy.
The production lines must be "hot." The US military decided to store critical machine tools after having munition manufacturing shortages in the Korean War, known as Plant Equipment Packages (PEPs). A retrospective report after the Cold War ended found that all the stored Army equipment was junk, but the Navy had kept their machines "energized" and utilized them part-time to create spare parts or training rounds. The excess capacity must be in active facilities even if they only run a few days a month or with skeleton crews during low demand.
Most subcomponents will be relatively standard parts that private firms produce. Contracts must pay for the right to extra capacity or provide payments that keep low-utilization lines afloat to ensure the entire supply chain can scale if needed.
A drone air force needs to create a helpful positive feedback loop. Short lifetimes and heavy training usage keep users sharp while reducing the logistical and maintenance load. The high turnover justifies constant manufacturing and helps support excess capacity. Continuous production and latent capacity help prepare for future conflicts.
Unifying Drone Strategy
Air power doctrine must change for smaller, cheaper drones to be useful outside rotor lift applications. And these drone models cannot act alone. Combined arms will be critical to maintain effectiveness. Applications like scouting, airborne early warning, tactical bombing, and missile/drone defense will be easier for drones to handle than air superiority or heavy bombing.
The priority for many countries after FPVs should be drones that operate over their lines. Weaknesses of drones like poor sensor capability, vulnerability to anti-aircraft artillery or tiny drone interceptors, limited range, and small payload are less critical over friendly territory. Providing alternatives to ground-based air defense missiles and frontline glide bombing at a fraction of the platform cost are massive improvements to the status quo.
Pushing past friendly lines requires integral scouting, more range, and counters for integrated air defenses. Survivability gravitates towards two poles, hiding or speed. Speed is expensive and only viable for high-end applications, like a hypersonic bomber. Low-end applications will gravitate to the hiding side. Loitering over enemy territory will be out of the question against competent militaries, and practicality will require mission-based sorties.
A mission would start with thousands of <10 kg scout drones spreading out across the area of operations. Their passive sensors make them hard to detect while also sipping power. The echelon of stealthy air-superiority drones would follow, creating a ragged picket line that makes it difficult for enemy interceptors to know where the line starts. These fighters would be loaded with air-to-air missiles and munitions to suppress ground-based air defenses (ship or shore). These formations create a bubble, complete with radar cluttering chaff, for the mission package to operate (typically tactical bomber drones or conventional aircraft). Stand-off range from glide bombs or small cruise missiles is critical to avoid anti-aircraft artillery and small drone interceptors.
Metrics like the kill-to-loss ratio become less relevant while others, such as cost-to-complete-mission, rise. The overall strategy is air denial and attrition rather than aggressive pursuit. Deterring enemy fighters from trying to intercept the mission is as much of a win as shooting them down. Progress is wearing down enemy airframes and crews, destroying infrastructure, denying air bases near the fighting, mining sea lanes, and disrupting logistics rather than directly destroying enemy systems.
The adaptations on each side will be extremely fast, so excess production capacity must back up any drone force instead of stockpiles that can quickly become worthless. Tomorrow's dog fights might happen at 100 km/hr a few thousand meters apart instead of Mach 2 and a 100-mile gap.
The net result of these drones would be to increase the cost of traditional deep penetration because they are so effective over friendly lines while reducing the cost of delivering munitions within 100-200 km of the front line falls because of cheap drones paired with air denial tactics.
The US Drone Force
Which branch should take on the drones?
The Air Force loathes low-performance aircraft and is skeptical of deleting human pilots. Its budget mostly goes towards capabilities that the drone air force isn't replacing, like deep strike, high-end fighters, or the nuclear umbrella. Deleting these platforms makes little sense when they still provide key capabilities (hedging!) and are in the phase where unit cost is falling. For those reasons, the Air Force is a poor choice to raise the drone force, and its job is to ensure its aircraft are protected from small drones when parked.
Thankfully, the US has four air forces to choose from, three of which already operate high-end aviation (Air Force, Navy, Marines).
-
Navy
The case for the Navy should be strong. Swarms of airborne early warning drones would be a viable replacement for the vulnerable E-2 Hawkeye. Air defense and superiority drones to provide a protective bubble around capital ships could be a necessary and valuable capability.
More fundamentally, the power of naval aviation peaked sometime in the mid-20th century. The increasing power and size of land-based aircraft made it expensive and difficult for ship-based aircraft to keep up. Stealth makes it even more challenging because it's already hard to have a top-end performance fighter that can take off and land on a carrier before adding it. Drones that have minimal sustainment needs and can fly off a converted container ship could revitalize naval power.
Will the US Navy do this? I'm somewhat doubtful.
-
Army
There is no question that the Army will have to adopt many of the rotor lift drones. It might need to acquire tactical bombers, air defense, and air superiority drones. The tactical bombers could fill in some rocket and tube artillery missions much more effectively. The cheap air-to-air drones have a cost advantage over the Army's ground-based air defense in stopping enemy drones and missiles. These needs will be especially pressing if the Air Force doesn't adopt them.
-
Marines
The Marines are trying to find a new identity after ~20 years of the Global War on Terror. They've ditched their tanks, bought anti-ship missiles, and adopted an "island hideout" strategy to be more relevant in East Asia. Air cover with a light footprint would theoretically fit that doctrine.
-
Drone Force?
Ukraine and Russia have created new branches for drone warfare. The primary reason to do this is that drones require extremely focused requirements/scope with fast iteration. The procurement processes for the traditional services are too bloated to do this well. A drone branch might manage the design, testing, procurement, and production of new drones with a relatively small personnel footprint. The drones could deploy to units of other branches with only a few operations soldiers from the drone units, almost like Navy corpsmen deploying with Marines.
The Basing Conundrum
Traditional air power's weaknesses are bases and sortie rate.
Long-range ballistic and cruise missiles already made protecting aircraft on the ground challenging. In East Asia, the US has abandoned any hope of using its strategically located bases on Okinawa in any Chinese conflict, choosing to fortify Guam and Tinian instead. Small drones add to the challenge by making every base vulnerable to special ops-style attacks.
Tactical aircraft can theoretically disperse and use smaller airstrips or highways, but larger support aircraft or bombers cannot. In some theaters, like East Asia, there aren't many alternative locations for tactical aircraft without changes to basing rights.
The US focuses on long-range aircraft and missiles instead of solving more local basing. Long-range is expensive, dings performance, and reduces productivity per aircraft.
The logic seems to assume that an opponent, like China, will only attack Taiwan, no other Asian nations will join, and the US will join the fight but not have a footprint on Taiwan. A more realistic plan would be figuring out how to operate from Taiwan, Luzon, and Japan. The US has operated aircraft from all three locations during previous Taiwan Strait conflicts, the Vietnam War, and the Korean War. Those aircraft can be a mix of traditional airframes and drones.
The Disruptive Potential of Space
I've previously written about space-based bombardment. These missions could provide a deep strike capability that drones aren't well suited for at a lower cost than stealth or hypersonic bombers.
SpaceX's Falcon 9 is likely at a cost and sortie rate that is strategically relevant. A back-of-the-envelope calculation suggests it can deliver 5x-10x more payload in a year than the US's current air-launched cruise missile production. Starship will only add to this. Other programs like Blue Origin's New Glenn and Rocket Lab's Neutron can add more capacity. Neutron can be more resilient to attack because of its lower dependence on launch tower facilities.
Pushing Drones Forward
In many ways, battlefield doctrine has not changed much since the 1940s. New weapons and advancements have driven design changes in tanks, ships, and aircraft but haven't altered their prominence. Rotor-lift drones fit in with leaps like anti-tank guided missiles, air-to-air missiles, and anti-ship missiles. Adding orbital strikes, a full drone air force, further development of ground drones, and artificial intelligence starts to stretch the comparison. Other factors, like more efficient ocean logistics, add more uncertainty. It is a time of uncertainty for the military balance of power.
Uncertainty requires hedging in a way that won't break the bank. Upgrades and life extensions for traditional platforms should win over starting new programs. Any new programs should be focused. Wariness should surround programs that don't fit the barbell. Do we need a $30 million subsonic fighter escort?
Drones will continue to be significant, and most effort should be on masses of low-performance aircraft optimized for new tactics. That requirement pushes heavily towards battery electric powertrains. These systems can only win by running traditional platforms into the ground with extreme sortie rate, logistical tail advantages, basing flexibility, and ease of production. Design, testing, and production capacity should be relatively inexpensive, enabling significant capability without betting the farm in the same way as a high-end program.
It seems likely that most people underestimate the likelihood of tail events, positive and negative. A war heavily involving naval and air forces has a decent probability of being extremely lopsided, especially with rapid technological change, given the historical record. Victories of a dominant nature can alter the balance of power. It pays to prepare.
Imagining the Drone Air Force
2025 January 20 Twitter Substack See all postsCan cheap, expendable drones complete every mission?
Features of Air Power
The side in control of the air tends to win. At a minimum, dominant air power is a massive force multiplier that allows the side wielding it to take significantly less casualties than its opponent. Aircraft can uniquely disrupt supply lines, command and control, and troop concentrations. The forces on the losing side must drastically alter their tactics to survive, limiting their ability to attack or defend.
Another feature is that air-to-air battles tend to be lopsided. It is more common to see 20:1 or 10:1 kill/loss ratios than even matches. For example, the F-15 has 104 kills and zero losses since entering service in 1976. The defining factors have been pilot quality, aircraft performance, weapon performance, and sensor capability (radar, airborne early warning aircraft, etc.).
US airpower was so dominant in the 20th century that most opponents focused on building ground-based anti-aircraft defenses. An arms race developed between these anti-aircraft missile batteries and ever more sophisticated aircraft, weapons, and tactics on the US side. Stealth to avoid detection, cruise missiles to avoid risking aircraft, and highly specialized tactics and weapons to defeat anti-aircraft batteries are an outgrowth of this competition.
Drones add a layer of complexity and prompt questions about how they fit into the traditional airpower paradigm.
What Drones Change
The most accurate paradigm for today's drones is that they are not a new category but dramatically reduce the cost of some existing functions. For example:
FPV Drones -> Attack Helicopters
Advocates of rotor aircraft thought they would dominate the battlefield in the 60s, 70s, and 80s, to the detriment of traditional armor. It didn't happen because helicopters are vulnerable to air defenses, including shoulder-fired missiles and anti-aircraft guns.
First Person View (FPV) kamikaze drones that cost <$1000 or slightly larger reusable drones are bringing this prediction back from the dead. They are still vulnerable to air defense, but it is irrelevant given their cost. Ground forces will need to make many adjustments, similar to when anti-tank guided missiles made WWII-style tanks obsolete in the 1960s and 1970s.
"Bomber Drones" -> Attack Helicopters pt. 2
Some missions call for slightly larger munitions than disposable FPVs can justify, and "bomber" drones that weigh around 25-50 kg and cost $10,000 fill the void. They mostly fly at night to increase survival rates and often use satellite communications, like StarLink, to avoid jamming. Missions are attacking parked vehicles, mining roads, and dropping grenades on infantry. These drones are much more powerful than FPVs and are worth the price if they can survive a few missions.
Recon Drones -> Scout Helicopters and Forward Air Control Aircraft
Scouting for artillery, ground attack aircraft, and attack helicopters has long been a scarce resource, even for the US military. Infantry and armor units still had to self-scout with limited visibility.
Small recon drones, often off-the-shelf commercial models, bring top-tier scouting down to the squad level. Their cost makes using them sustainable, while many large drones, like the US Predator, are obsolete in high-intensity battles because of their price and vulnerability to air defenses.
One-Way Attack Drones -> Cruise Missiles
Cruise missiles have a unique ability to attack heavily defended targets in the opponent's rear, but their price limits their number.
Propeller-powered one-way attack drones can cost as little as $50,000 instead of $1+ million, increasing volume. The overall impact has been much more muted than FPV and recon drones because these drones are so easy to shoot down and have small payloads that limit what targets they can be effective against. They travel slowly, roughly the same as a car on the interstate, to meet cost goals and extend range. Their utility plummets once the opponent adapts to shoot them down with cheap weapons, like guns on trucks and helicopters, cheap interceptor drones, or electronic warfare. The drones can still provide net benefits if they temporarily overwhelm air defenses, force the enemy to expend significant organizational resources to counter them, or the targets are valuable enough.
Interceptor Drones -> Man Portable Anti-Aircraft Missiles
Militaries developed man-portable anti-aircraft missiles to counter helicopters and low-flying aircraft, but they are much too expensive and complex to counter drones.
Instead, small racing-style drones that cost no more than a few thousand dollars ram or explode near targets. Their prey is primarily more expensive attack drones and higher-tier recon drones that cost $30,000-$200,000.
There are some experiments with drones carrying shotguns and other air-to-air weaponry to deal with the smallest FPV and recon drones. Time will tell if these are viable.
Some categories haven't seen drones encroach on their territory, namely air-superiority fighters or bombers.
The main battlefield effects are that low-level rotor-based attack is effective, detection of movement is much more likely, integrated air defense systems have to handle an order of magnitude more bogies, and slow speed, low altitude aircraft are becoming vulnerable to small interceptors.
Analyzing Trends
The rapid development of these categories tells us several things about the future of drones:
Barbell Procurement Strategy
The battlefield is so hostile that drones must be cheap enough to be expendable or capable enough to avoid all air defenses. The somewhat fancy $100,000 recon drone is probably in no-man's-land. Large drones without sophisticated countermeasures, like the US Global Hawk or Predator/Reaper family, are obsolete outside the most permissive airspace. Even drones that were considered cheap before the war in Ukraine, like the Turkish TB-2, have been sent to the scrap heap.
One of the only viable(?) large drones currently in use is the pricey US RQ-180 because of its size and modern stealth features. Traditional cruise missiles also continue to be viable for deep strikes.
Small Eats Large
Drones aren't automatically cheaper than legacy systems like helicopters or strategic reconnaissance platforms. Radical reduction in size and complexity is the best way to achieve this.
Better electronics and cameras have allowed recon drones with mass measured in grams. Or a shaped charge driven by an FPV drone into the weakest part of a vehicle's armor can be much smaller than traditional anti-tank missile warheads.
Battery-electric powertrains can shrink much more than engines can, and these drones have disrupted short-range, low-speed categories much more than long-range or high-powered missions.
The success rate of these drones is often low, between 10%-50%, and many targets need multiple hits. However, the low cost of small drones means the math is favorable and similar to artillery shells.
Single Function Dominates
Drones can only be small and cheap if they are highly specialized for one task. Examples include anti-vehicle, anti-personnel, high-value targets ~30 km behind enemy lines, hitting enemy drones, dropping mines or supplies, etc. Many of these categories even have further specialization within them.
Paths with Faster Iteration Win
Things change fast since small drones are a relatively new technology. Pathways that allow quick adjustments can outcompete slow paths. Small and single-function platforms can increase iteration speed.
The Drone Air Force
Drones have changed the battlefield by supercharging army rotor aviation functions and organic air defense. They only partially substitute for traditional air power functions like close air support or air defense/superiority.
A true drone air force must keep the lessons of the barbell, small, single function, fast iteration, and proper powertrain in mind to revolutionize air power. Otherwise, it will be a traditional aircraft without a pilot. The designs must provide cheap precision, be reliable, and be easy to mass-produce to stay relevant in high-intensity conflict.
The single function requirement provides the guide for what models are needed. The designs assume efficient and tightly integrated designs - high payload/battery fraction, high lift-over-drag ratio, etc. The SpaceX/Tesla treatment rather than the catalog engineering special.
Airborne Early Warning (AEW)
Radar signal strength diminishes rapidly, requiring powerful radars to see long distances. Those radars are too bulky and expensive to put on each fighter. Air forces convert civilian airliners to radar aircraft instead. Opponents can usually detect radar signals at double their range, so it's helpful for fighters to keep their radars off to conceal their positions and get directions from the airborne early warning (AEW) aircraft.
These legacy AEW platforms are slow, expensive, and obsolete in high-intensity warfare. Ukraine has knocked out several of Russia's without an effective air force. Long-range missiles with final radar guidance are the kiss of death for planes like the US E-2 or E-3.
The US Air Force wants to distribute this capability in $10-$30 million drones that fly as "loyal wingmen" to manned stealth fighters. The procurement cost for the hundreds of these drones needed to fulfill the same function as 24/7 coverage from traditional AEW platforms would be similar to buying E-3 replacements.
But the drone success algorithm might produce something better. Passive sensors are lighter, cheaper, and use less power, but they also have short detection ranges and are more sensitive to weather. It might require 5000-10,000 drones to replace one 24/7 circuit. However, it is an application that can work for electric batteries, meaning the total cost will probably be 1/100th of jet drones or traditional AEW platforms.
The drones would be fixed-wing, powered gliders weighing a few kilograms, optimized to circle a small area for 6-10 hours. They would hunt for enemy aircraft or missiles using cameras, infrared, and microphones and send bursts of information if their onboard processing notices anything interesting. The cost would be <$5000/unit to absorb losses from interceptor drones. Their sensors, batteries, and software will improve much faster than the multi-billion dollar development projects that are the alternative.
Air/Missile Defense
Air defense today usually means stopping incoming cruise missiles and drones rather than dogfighting. Tactical fighters like the F-16 are capable but overkill for the role. Aircraft are valuable in this role because their mobility allows them to deploy more places and get close to their targets. Close engagements require less powerful radars and smaller, cheaper missiles than a ground-based battery covering a large area.
The ideal munition to base the vehicle around is probably the APKWS rocket or another comparable short-range missile/rocket like the Zuni. It is mass-produced, cheap, and accurate as long as it's within a few kilometers of the bandit. It is proven against larger drones in Ukraine and successfully tested against cruise missile-type targets.
An off-the-shelf option is putting missiles or rockets on the Kratos XQ-58, which costs $5-$10 million kitted out. Its range and payload are overkill when opponents plan missile and drone salvos to have their weapons converge from multiple angles, limiting how many one aircraft can intercept.
The "safe" answer for propulsion is jet engines. High subsonic speed is the minimum to counter cruise missiles with a decent interception rate. Liquid fuel enhances the loiter time and increases the area the aircraft can defend. Anduril's Roadrunner is an example of what the aircraft might look like. They sell them for ~$500,000 each, including some auxiliary systems.
A battery-electric drone architecture would be an exercise in brute force. The fastest electric drone is around 300 mph and not quick enough for conventional intercepts. Aircraft power increases at the cube of velocity, so faster drones can end up on the wrong side of the cost curve. Instead, the strategy would be multiple "picket" lines of slow, tightly spaced drones. The APKWS only weighs 15 kg, and available radar models with enough range to guide small rockets/missiles weigh <20 kg . The airframe and propulsion can be flimsy and cheap because of the small payload and slow speed. The main challenge would be making the cost of the avionics package reasonable. These would exclusively operate over friendly territory to maintain survivability.
The algorithm would prefer the brute force, slow, and cheap drone strategy. Over time, better batteries and motors should improve performance while avionics costs fall. The logistics tail and basing requirements would be minimal compared to jets, and the drone would ideally be cheap enough to keep adding picket lines as necessary. It is a rapidly scalable air defense that is more cost-effective than fixed missile batteries and more challenging for opponents to game plan due to the inherent mobility.
Air Superiority
The current thinking is that modern air-to-air engagements between top-tier air forces begin like a Wild West duel. Each side rips off salvos of long-range missiles, hoping to disable their opponents first. Engagement distances might remain long with no visual range combat. One side in any hypothetical engagement almost certainly has an advantage in front-line stealth fighters, pilots, and missiles that would end the air battle relatively quickly.
The requirements for an air superiority fighter are locating opponent aircraft and firing weapons that can get a hit. The point of fighters like the F-22 is to make this difficult. They are hard to detect, challenging to get a missile lock on, have exceptional agility, and can push over Mach 2. The US air-to-air missiles, which presumably have a chance at a target like this, range from $400,000 to several million dollars. And it takes a skilled pilot in a capable aircraft to get in position for a high-quality shot.
The US Air Force's air superiority drone strategy is relatively capable models ($10-$30 million) that carry more missiles and scout for manned fighters.
A risk with our cheap drone algorithm is that stealth fighters can load up with dozens of cheap air-to-air missiles that easily handle slow, visible drones. The fighters can rip apart drone swarms without any drone being able to get a firing solution because the drones only have short-range radars, and swarm intelligence often provides qualitative instead of quantitative vectors. The cost could be upside down since these won't be $5000 drones, plus each one carries a missile costing >$400,000.
A cheap, battery-electric air superiority drone has to cheat. A small, slow drone is easier to make stealthy than a supersonic stealth fighter. There are typically tradeoffs between stealth and maneuverability, like not having a tail, but losing maneuverability is fine here. The small size helps further reduce the radar cross-section and visibility. The thermal signature would be minimal with such an efficient powertrain. And radar-absorbing paint doesn't have to last long or be the best. Decoys and clutter-like chaff will be much more effective when the drones are slow-moving. A drone like this could also have a very damage-resistant design compared to a fighter that is mostly fuel and a massive engine spinning at thousands of RPM. The low signature would hinder the opponent's ability to locate and kill the drones with simple, cheap missiles, while durability increases the opponent's required missile size/amount.
Flying at very slow speeds would also provide the range and loitering time to patrol forward of friendly lines, though it would not be able to operate from rear bases. Engagements would be extremely close range and facilitated by the sheer number of drones spaced every few kilometers. Visible light, IR, and synthetic aperture radar sensors would have a chance of detecting opponent stealth fighters at short range. Air-to-air missiles can be less sophisticated because they need minimal onboard guidance for close-in engagements, and the target aircraft would only have a few seconds to maneuver.
The strategy would be more like "air denial" than air superiority. American-style air dominance is aggressive, chasing down bogies while accompanying aircraft attack air bases. A slow-moving cloud of drones is easy to avoid but creates a no-go zone.
There is a good chance the low-end drone strategy doesn't work in air-to-air combat, given the history and how extreme the evolutionary pressure will be to locate and out-range these no-man's-land cost drones with cheap munitions. Achieving one kill for 100 or 1000 losses is not given against modern fighters, nor is one kill for 10 to 100 losses against low-end jet drones. The evolutionary successful strategy could be an AI pilot flying a $200 million fighter. Cheap air denial drones could still be helpful in creating temporary bubbles for other aircraft to work. It is possible that traditional air superiority aircraft lose because they are vulnerable on the ground and dependent on support aircraft (tankers, airborne early warning) rather than any deficiencies in the air.
Suppression and Destruction of Air Defense
The early days of neutralizing surface-to-air missile batteries were harrowing work. Attacking aircraft would fly right above the treetops and pop up to fire a radiation-seeking missile or drop cluster munitions. The sites were simple to disable because missiles needed central guidance to the targeted aircraft. Radars were single-point failures. The evolution has been toward more capable surface-to-air missiles and independent subunits, making them much more resilient. Attackers now use longer-range, fast anti-radiation missiles with multi-mode seekers to track down vehicles that cut and run after seeing a missile launch. These methods still seem adequate and effective for high-end, non-Western air defenses.
Towed anti-aircraft artillery is one obvious counter to slow drones. These guns are cheap, easy to repair, don't require precise target location, and can handle high volumes of drones. Drone formations would need a new class of cheap missile/loitering munition for these weapons. They probably look like Anduril's Altius series of drones, but 10x cheaper.
Tactical Bombing
Glide bombs have become a staple of US aviation and one of the most effective weapons in the Ukrainian war. They are cheap, accurate, and provide enough stand-off range to increase the survivability of the launching aircraft. Their size (250-2000 lb) packs more punch than small drones or tube artillery. Air superiority allows tactical bombers wielding the bombs to wreak havoc behind enemy lines. They can pound enemy positions and forward command posts near the line of contact if enemy airspace is well-defended.
The off-the-shelf option for a drone tactical bomber is, again, the Kratos XQ-58, which can carry four small diameter bombs with a several thousand-kilometer range.
The algorithm would suggest picking a smallish glide bomb, basing the airframe design around carrying a single bomb, and checking if an electric powertrain can meet the range requirements.
The output looks something like an aircraft that weighs 300 kilograms and carries a 120-kilogram small diameter bomb. A battery-electric powertrain could give it 150-300 kilometers of combat radius, plus 20-100 km from the bomb's gliding. Flight speed would be relatively slow, but it would accelerate briefly at release to give the bomb more velocity or at takeoff to use short fields/roads. The battery would be about 1/5 the capacity of a Model 3, and the drone could fit in the bed of an F-150 if the wings fold in. The range is relatively conservative because such a drone could use battery cells with no reserve and minimal fire safety systems. The vehicle might barely cost more than the bomb it carries($50,000), so some losses are acceptable, but the general flight path is low risk because of the stand-off from the glide bomb.
That capability would be relevant in Ukraine, Syria, Taiwan, and other possible theaters. The development cost could be low, so it would not be a loss if the design became obsolete. And US allies could operate these aircraft. Moving tactical bombing to small, electric platforms that don't need airfields simplifies the logistical tail, reduces reliance on vulnerable and politically contentious regional bases, and provides organic tactical bombing without constant overhead coverage from traditional airframes.
Sea Mine Laying
Bomber-dropped mines were one of the crucial blows against the Japanese in WWII and could have similar impacts against some potential US opponents. The trouble is dropping them in contested airspace with enough volume to matter. One recent innovation has been to turn glide bombs into mines, providing stand-off range for aircraft.
Defensive mines in the shallowest water and harbors, like for Taiwan or the Philippines, are easy, along with compact regions like the Persian Gulf or Baltic Sea. The solution looks like the tactical bomber.
Slightly larger battery-electric drones can mine most of the Chinese coast north of Taiwan, flying from South Korea, Taiwan, or Japan.
Southern China is much harder to access and would likely be handled by sophisticated, multi-platform air operations or other means, like submarines.
Strategic Bomber
Mass and range matter with bombers, making a cheap drone paradigm challenging. Bombers are most important in contested airspace because tactical aircraft can handle most attacks once the "wild weasels" degrade air defenses. Performance is required. The US B-21 meets the need.
One logical alternative to stealth bombers is concepts like the US Air Force's Rapid Dragon, which allows cargo planes, like the C-130, to launch cruise missiles. The C-130 is still in production after decades, has a good payload and range, is cost-effective to buy and operate, and can take off from rough airfields. The plane is a drone carrier with cruise missiles being the drones.
A cutting-edge replacement would be reusable hypersonic drones. These aircraft would use speed to avoid enemy defenses. Munitions can be simpler and cheaper than cruise missiles because the glide distance of a bomb launched at Mach 5 from 70,000 feet can be hundreds of kilometers. The game-breaking speed that outruns missiles and increases standoff range allows a hypersonic drone to stay within the algorithm. Engineers can focus on making it fly fast instead of traditional fighters that require sophisticated engineering across many domains to remain alive for more than a few minutes in hostile airspace. An aircraft like this would be disruptive by allowing attacks in the enemy rear before gaining air superiority and potentially having a much higher sortie rate than slow stealth bombers.
Munition Innovation
Today's munitions are optimized for the pre-smartphone era, often having expensive electronics built to military specifications that aren't commercial-off-the-shelf friendly. Guidance kits for rockets and bombs turn $500-$3000 munitions into $30,000-$50,000 ones. That is a steal compared to million-dollar cruise missiles but ridiculous compared to what could be. Incorporating computer vision and other cheap sensors seems like low-hanging fruit to lower cost.
There also seems to be some convergence on cruise missiles/drones that are a tenth the size of traditional cruise missiles with similar speed and less range. These require new, small jet engines, and the diversity in effort is remarkable. Several groups are dusting off the German WWII V-1 pulse jet. Others like Kratos or Ukrainian groups are leveraging their years of making cheap engines for target drones or other lower-tier applications. And some, like Anduril, are developing clean-sheet designs that they hope to use in many applications.
New classes of bombs, missiles, and sea mines optimized for drones would amplify effectiveness.
Manufacturing and Logistics
The creation and operation of a drone air force differs from modern air forces.
Operations
The slow, electric drones have massive advantages in operating footprint and costs, with that impact reverberating down the logistics tail:
Fuel
Fuel is one of the biggest concerns for modern militaries. The US military often assumes a fuel cost of hundreds or thousands of dollars per gallon to deliver to war zones for planning. The volume of demand in a high-intensity conflict could reach the level of economies like Japan. Aircraft are often the largest fuel consumers.
These drones require a fraction of the energy of high-performance aircraft and can often use electricity instead of fuel. The AEW scout battery pack would only be a few pounds and could charge with a tiny solar panel. Models like the tactical bomber, air-defense fighter, or air superiority fighter have batteries small enough to swap by hand. A few standard-size solar panels could provide enough juice for one sortie per day and don't require vulnerable centralized infrastructure.
Infrastructure
An aircraft's weight and stall speed plays a large part in determining runway length and quality. Small, slow drones need minimal airstrips, if they need them at all. There is no need for traditional air bases.
Parts/Maintenance
The US military prefers "module-based" maintenance. Techs change an entire radar module instead of diagnosing and fixing a certain subcomponent to reduce labor hours and the number of parts in stock.
Many drones would cost as much as a typical module, and there would be no reason to bother with parts or repairs. The need for techs and parts management would be minimal.
Battery-electric powertrains are reliable compared to jet engines and should be able to fly hundreds or thousands of hours before replacement without maintenance.
Training
Fighter pilots are the most valuable rank-adjusted human capital in any military. One great pilot can make a meaningful difference in an entire war by helping to clear the skies. Selection is intense, and training is very slow. Simulators help, but learning to fly a $100+ million fighter jet doesn't happen overnight.
AI pilots take more effort to train initially but can replicate as needed. The burden for any human pilot/manager will be lower given the narrower mission of drones than multi-role fighters. The low cost of the platforms means both AI and humans can train constantly on real aircraft instead of using simulators. Battles can be live instead of simulated without endangering human pilots, improving the quality of training.
Sortie Rate
Most fighters need full crews to turn the aircraft around and keep it flying. Each airframe only has so many hours without full refits. Many aircraft struggle to fly one sortie per day. A low-maintenance, battery-electric drone with swappable batteries could fly 20-22 hours each day.
The sortie challenge would be especially beneficial for countries like Taiwan. China constantly flies fighters at the edge of Taiwan's air defense identification zone, which forces Taiwan to send fighters to intercept them, wearing down airframes and pilots. A constant picket of drones would negate this strategy.
Shipping
Munitions, especially bombs, are the biggest logistical challenge after fuel. Manned aircraft tend to drop large bombs that are overkill because of their limited sortie rates and the risk each mission entails. Drones with high sortie rates can use small bombs that make the drone more practical and reduce total tonnage dropped as each target gets the appropriate amount instead of a truck getting vaporized by a 2000 lb bomb.
Battery-electric drones are a dream from an operations and logistic standpoint. High sortie rate, no fuel deliveries, low cost per flight hour, no maintenance tail, and reduced munition tonnage are all features.
Manufacturing
Today's militaries build aircraft in small numbers (<1000), with production runs lasting a decade or two. Only the most popular models, like the F-16 or the C-130, stay in production and reach thousands of units. Airframes last decades and undergo upgrades - new radars, wings, etc. - to extend their life. The upfront expense for new aircraft is eye-watering, and the fixed cost is high because of the low production numbers.
A drone air force would be the opposite. The development cost for single-purpose drones with battery electric power trains would be relatively low. Production volume would be extremely high. And the lifetimes of the drones would be short with no upgrades.
Production needs to absorb lessons from other high-volume programs, like munitions. Demand skyrockets during wars, but production lags because factories take time to come online. The two main strategies to deal with this are stockpiles and underutilized factories.
Stockpiles are common for artillery shells where manufacturing equipment is expensive, designs rarely change, and the shells have long shelf lives. Drones change fast, and production is more akin to light manufacturing, meaning extra factory capacity is a better strategy.
The production lines must be "hot." The US military decided to store critical machine tools after having munition manufacturing shortages in the Korean War, known as Plant Equipment Packages (PEPs). A retrospective report after the Cold War ended found that all the stored Army equipment was junk, but the Navy had kept their machines "energized" and utilized them part-time to create spare parts or training rounds. The excess capacity must be in active facilities even if they only run a few days a month or with skeleton crews during low demand.
Most subcomponents will be relatively standard parts that private firms produce. Contracts must pay for the right to extra capacity or provide payments that keep low-utilization lines afloat to ensure the entire supply chain can scale if needed.
A drone air force needs to create a helpful positive feedback loop. Short lifetimes and heavy training usage keep users sharp while reducing the logistical and maintenance load. The high turnover justifies constant manufacturing and helps support excess capacity. Continuous production and latent capacity help prepare for future conflicts.
Unifying Drone Strategy
Air power doctrine must change for smaller, cheaper drones to be useful outside rotor lift applications. And these drone models cannot act alone. Combined arms will be critical to maintain effectiveness. Applications like scouting, airborne early warning, tactical bombing, and missile/drone defense will be easier for drones to handle than air superiority or heavy bombing.
The priority for many countries after FPVs should be drones that operate over their lines. Weaknesses of drones like poor sensor capability, vulnerability to anti-aircraft artillery or tiny drone interceptors, limited range, and small payload are less critical over friendly territory. Providing alternatives to ground-based air defense missiles and frontline glide bombing at a fraction of the platform cost are massive improvements to the status quo.
Pushing past friendly lines requires integral scouting, more range, and counters for integrated air defenses. Survivability gravitates towards two poles, hiding or speed. Speed is expensive and only viable for high-end applications, like a hypersonic bomber. Low-end applications will gravitate to the hiding side. Loitering over enemy territory will be out of the question against competent militaries, and practicality will require mission-based sorties.
A mission would start with thousands of <10 kg scout drones spreading out across the area of operations. Their passive sensors make them hard to detect while also sipping power. The echelon of stealthy air-superiority drones would follow, creating a ragged picket line that makes it difficult for enemy interceptors to know where the line starts. These fighters would be loaded with air-to-air missiles and munitions to suppress ground-based air defenses (ship or shore). These formations create a bubble, complete with radar cluttering chaff, for the mission package to operate (typically tactical bomber drones or conventional aircraft). Stand-off range from glide bombs or small cruise missiles is critical to avoid anti-aircraft artillery and small drone interceptors.
Metrics like the kill-to-loss ratio become less relevant while others, such as cost-to-complete-mission, rise. The overall strategy is air denial and attrition rather than aggressive pursuit. Deterring enemy fighters from trying to intercept the mission is as much of a win as shooting them down. Progress is wearing down enemy airframes and crews, destroying infrastructure, denying air bases near the fighting, mining sea lanes, and disrupting logistics rather than directly destroying enemy systems.
The adaptations on each side will be extremely fast, so excess production capacity must back up any drone force instead of stockpiles that can quickly become worthless. Tomorrow's dog fights might happen at 100 km/hr a few thousand meters apart instead of Mach 2 and a 100-mile gap.
The net result of these drones would be to increase the cost of traditional deep penetration because they are so effective over friendly lines while reducing the cost of delivering munitions within 100-200 km of the front line falls because of cheap drones paired with air denial tactics.
The US Drone Force
Which branch should take on the drones?
The Air Force loathes low-performance aircraft and is skeptical of deleting human pilots. Its budget mostly goes towards capabilities that the drone air force isn't replacing, like deep strike, high-end fighters, or the nuclear umbrella. Deleting these platforms makes little sense when they still provide key capabilities (hedging!) and are in the phase where unit cost is falling. For those reasons, the Air Force is a poor choice to raise the drone force, and its job is to ensure its aircraft are protected from small drones when parked.
Thankfully, the US has four air forces to choose from, three of which already operate high-end aviation (Air Force, Navy, Marines).
Navy
The case for the Navy should be strong. Swarms of airborne early warning drones would be a viable replacement for the vulnerable E-2 Hawkeye. Air defense and superiority drones to provide a protective bubble around capital ships could be a necessary and valuable capability.
More fundamentally, the power of naval aviation peaked sometime in the mid-20th century. The increasing power and size of land-based aircraft made it expensive and difficult for ship-based aircraft to keep up. Stealth makes it even more challenging because it's already hard to have a top-end performance fighter that can take off and land on a carrier before adding it. Drones that have minimal sustainment needs and can fly off a converted container ship could revitalize naval power.
Will the US Navy do this? I'm somewhat doubtful.
Army
There is no question that the Army will have to adopt many of the rotor lift drones. It might need to acquire tactical bombers, air defense, and air superiority drones. The tactical bombers could fill in some rocket and tube artillery missions much more effectively. The cheap air-to-air drones have a cost advantage over the Army's ground-based air defense in stopping enemy drones and missiles. These needs will be especially pressing if the Air Force doesn't adopt them.
Marines
The Marines are trying to find a new identity after ~20 years of the Global War on Terror. They've ditched their tanks, bought anti-ship missiles, and adopted an "island hideout" strategy to be more relevant in East Asia. Air cover with a light footprint would theoretically fit that doctrine.
Drone Force?
Ukraine and Russia have created new branches for drone warfare. The primary reason to do this is that drones require extremely focused requirements/scope with fast iteration. The procurement processes for the traditional services are too bloated to do this well. A drone branch might manage the design, testing, procurement, and production of new drones with a relatively small personnel footprint. The drones could deploy to units of other branches with only a few operations soldiers from the drone units, almost like Navy corpsmen deploying with Marines.
The Basing Conundrum
Traditional air power's weaknesses are bases and sortie rate.
Long-range ballistic and cruise missiles already made protecting aircraft on the ground challenging. In East Asia, the US has abandoned any hope of using its strategically located bases on Okinawa in any Chinese conflict, choosing to fortify Guam and Tinian instead. Small drones add to the challenge by making every base vulnerable to special ops-style attacks.
Tactical aircraft can theoretically disperse and use smaller airstrips or highways, but larger support aircraft or bombers cannot. In some theaters, like East Asia, there aren't many alternative locations for tactical aircraft without changes to basing rights.
The US focuses on long-range aircraft and missiles instead of solving more local basing. Long-range is expensive, dings performance, and reduces productivity per aircraft.
The logic seems to assume that an opponent, like China, will only attack Taiwan, no other Asian nations will join, and the US will join the fight but not have a footprint on Taiwan. A more realistic plan would be figuring out how to operate from Taiwan, Luzon, and Japan. The US has operated aircraft from all three locations during previous Taiwan Strait conflicts, the Vietnam War, and the Korean War. Those aircraft can be a mix of traditional airframes and drones.
The Disruptive Potential of Space
I've previously written about space-based bombardment. These missions could provide a deep strike capability that drones aren't well suited for at a lower cost than stealth or hypersonic bombers.
SpaceX's Falcon 9 is likely at a cost and sortie rate that is strategically relevant. A back-of-the-envelope calculation suggests it can deliver 5x-10x more payload in a year than the US's current air-launched cruise missile production. Starship will only add to this. Other programs like Blue Origin's New Glenn and Rocket Lab's Neutron can add more capacity. Neutron can be more resilient to attack because of its lower dependence on launch tower facilities.
Pushing Drones Forward
In many ways, battlefield doctrine has not changed much since the 1940s. New weapons and advancements have driven design changes in tanks, ships, and aircraft but haven't altered their prominence. Rotor-lift drones fit in with leaps like anti-tank guided missiles, air-to-air missiles, and anti-ship missiles. Adding orbital strikes, a full drone air force, further development of ground drones, and artificial intelligence starts to stretch the comparison. Other factors, like more efficient ocean logistics, add more uncertainty. It is a time of uncertainty for the military balance of power.
Uncertainty requires hedging in a way that won't break the bank. Upgrades and life extensions for traditional platforms should win over starting new programs. Any new programs should be focused. Wariness should surround programs that don't fit the barbell. Do we need a $30 million subsonic fighter escort?
Drones will continue to be significant, and most effort should be on masses of low-performance aircraft optimized for new tactics. That requirement pushes heavily towards battery electric powertrains. These systems can only win by running traditional platforms into the ground with extreme sortie rate, logistical tail advantages, basing flexibility, and ease of production. Design, testing, and production capacity should be relatively inexpensive, enabling significant capability without betting the farm in the same way as a high-end program.
It seems likely that most people underestimate the likelihood of tail events, positive and negative. A war heavily involving naval and air forces has a decent probability of being extremely lopsided, especially with rapid technological change, given the historical record. Victories of a dominant nature can alter the balance of power. It pays to prepare.