Aerospace technology is not evolving in a straight line right now. It is splitting in two directions at once, and understanding that split is the key to understanding almost everything happening in the field in 2026.
On one side sits an old, familiar trajectory: exquisite, expensive, low-volume platforms that push the absolute limits of speed, stealth, and sensor performance. The Boeing F-47, the Global Combat Air Programme’s Tempest, Lockheed Martin’s hypersonic glide bodies. These are aircraft and weapons that cost hundreds of millions or billions of dollars, take a decade or more to field, and will number in the dozens or low hundreds when they finally reach service.
On the other side sits something genuinely new: cheap, numerous, software-defined systems built explicitly to be produced fast, lost without consequence, and replaced from a digital file rather than a factory floor. Autonomous rotorcraft flying as expendable wingmen. Guided 50mm cannon rounds priced to out-attrit drone swarms. Robotically formed aircraft panels built one press at a time because the original supplier vanished a decade ago.
Neither trajectory is winning. Both are accelerating simultaneously, and the connective tissue between them, the software, the sensors, and increasingly the orbital architecture watching over all of it, is where the real story of the next decade is being written. This piece pulls together six forces currently reshaping aerospace technology and examines how they interact, where they are heading, and what still stands in the way.
1. The Crewed Fighter’s Uncertain Century
The first and perhaps most counterintuitive trend to understand is that the crewed fighter jet is not disappearing. Despite two decades of predictions that autonomous systems would render piloted combat aircraft obsolete, every major air power currently developing a next-generation fighter is building one with a cockpit in it.
The United States awarded Boeing a contract worth roughly $20 billion in March 2025 to build the F-47, the crewed centrepiece of the Air Force’s Next Generation Air Dominance (NGAD) programme and successor to the F-22 Raptor. The aircraft is on track for a first flight in 2028, and it is being designed less as a dogfighter than as what industry observers have taken to calling a “quarterback”: a stealthy, long-range aircraft with a combat radius exceeding 1,000 nautical miles, built to command a formation of autonomous Collaborative Combat Aircraft (CCA) flying ahead of and alongside it into contested airspace. The Air Force’s stated ambition is to field more than 1,000 of these CCAs, roughly two for every F-47, meaning the aircraft’s real combat value lies as much in what it commands as in what it alone can do.
Britain, Italy, and Japan are pursuing the equivalent ambition through the Global Combat Air Programme (GCAP), built around the UK’s Tempest concept and Japan’s Mitsubishi F-X effort, merged under a joint venture called Edgewing. The programme cleared a major milestone on July 3, 2026, when the three governments signed a £4.6 billion contract pushing the aircraft into full-scale detailed design, with entry into service targeted for 2035 and a demonstrator aircraft expected to fly in 2027. Canada joined the effort as an observer on July 21, 2026, a sign of the pull a sovereign, non-American sixth-generation fighter programme now exerts on allies increasingly wary of depending entirely on U.S. platforms.
Not every next-generation fighter programme has fared as well. Future Military Technologies covered the collapse of the Franco-German Future Combat Air System earlier this year, a reminder that international sixth-generation partnerships carry real political and industrial risk alongside their cost-sharing benefits. Meanwhile, Russia’s Su-75 Checkmate has only just reached the physical prototype construction stage after five years of delay, illustrating how difficult it is for even a major aerospace power to execute a clean-sheet fighter programme without an alliance to share the burden.
The analytical point worth drawing out here is this: even in an era of accelerating autonomy, no major power has concluded that crewed combat aircraft are finished. What has changed is the job description. The crewed fighter is increasingly a sensor-fused, long-range command node for a much larger uncrewed formation, not simply a faster, stealthier version of the aircraft it replaces.
2. The Rise of the Uncrewed Formation
If the crewed fighter’s job is changing, the reason is the second major force reshaping aerospace: the sheer scale at which uncrewed systems are now being planned into force structure, not as auxiliary tools but as the numerical bulk of future combat power.
The clearest illustration is Anduril’s Thunder, which Future Military Technologies examined in detail in July, an autonomous tiltrotor attack aircraft co-developed with Archer Aviation and designed specifically to team with crewed helicopters like the Apache and the incoming Bell MV-75 Cheyenne. Anduril’s own figures for the concept are telling: pairing three Thunder aircraft with a single Apache triples the munitions a Combat Aviation Brigade can bring to a fight without adding a single additional pilot. That is the uncrewed formation thesis in one sentence, mass multiplied without proportional risk to human crews or proportional growth in acquisition cost.
The same logic is playing out on the ground and at sea. The U.S. Marine Corps’ contract with Overland AI for autonomous logistics vehicles supporting its counter-drone air defence batteries reflects the same pattern in miniature: a small, modest contract by defence-industry standards, but the first time a ground autonomy company has held a production prime contract with the U.S. military, exactly the kind of institutional threshold-crossing that tends to precede much larger commitments.
What ties these programmes together is not the hardware but the software underneath it. Anduril’s Lattice for Mission Autonomy platform, the same software backbone behind Thunder, also underpins the sensor fusion and formation logic in the company’s other autonomous systems, and it is a core component of the British Army’s ASGARD digital targeting network, which Future Military Technologies covered in July. That overlap is not a coincidence. The genuinely scarce and valuable resource in the uncrewed formation era is not the airframe, ground vehicle, or hull, all of which are becoming increasingly commoditised, but the autonomy stack that lets a human operator issue intent rather than fly stick-and-rudder, and lets dozens of platforms deconflict, share sensor data, and execute coordinated tasking at machine speed.
3. Propulsion’s Quiet Revolution, and Its Growing Bottleneck
Autonomy and stealth tend to dominate aerospace headlines, but propulsion is where some of the most consequential engineering is currently happening, and where the schedule risk for nearly every next-generation programme is concentrated.
The centrepiece of American fighter propulsion development is the Next Generation Adaptive Propulsion (NGAP) programme, competing designs from GE Aerospace (the XA102) and Pratt & Whitney (the XA103) for a three-stream adaptive cycle engine intended to power the F-47. Unlike a conventional turbofan, an adaptive cycle engine can shift its internal airflow configuration in flight, favouring raw thrust in a high-speed engagement and fuel efficiency during a long-range cruise, and Air Force officials expect roughly 30 percent more range and double the thermal management capacity compared with current engines, a genuinely significant figure given how much heat sixth-generation sensors, computers, and electronic warfare systems generate. But the programme has slipped. Both companies only completed assembly readiness reviews for their prototype engines in May 2026, and Air Force officials confirmed on July 27, 2026, that the F-47 will fly its first flight in 2028 using an interim propulsion arrangement, roughly two years before the NGAP engine intended for the operational aircraft is ready for integration. Frank Kendall, the former Air Force Secretary, has called the earlier termination of a related adaptive-engine effort under the F-35 programme his greatest budgetary regret, a telling admission of just how highly the service values this technology and how costly it has been to fund inconsistently.
Hypersonic propulsion tells a parallel story of ambition outrunning schedule, though from a different starting point. Lockheed Martin’s Next Generation Glide Body (NXGB), unveiled in June 2026, represents an explicit pivot away from treating hypersonic weapons purely as a performance problem and toward treating them as a manufacturing problem, built around a wedge-shaped, lifting-body design intended to be genuinely producible at scale rather than hand-built in small numbers, a direct response to the U.S. Army’s own decision to pivot away from further procurement of the currently fielded Dark Eagle system in favour of cheaper alternatives.
Elsewhere in propulsion, the hybrid-electric powertrain behind Anduril’s Thunder, and the Optimum-Speed Tiltrotor technology that varies rotor RPM to balance efficiency against acoustic signature, points toward a third propulsion frontier distinct from both fighter jets and hypersonic weapons: electrified vertical lift, developed jointly with the commercial eVTOL sector and cross-pollinating military and civilian aviation in a way that has not happened at this scale before.
The pattern across all three areas is consistent. Propulsion, not airframe design, sensors, or even software, is increasingly the pacing item that determines when a next-generation aerospace system actually reaches the field. Airframes can be shaped in a digital wind tunnel faster than an adaptive-cycle engine can be matured and certified.
4. Software as the New Airframe
The fourth force is the one that cuts across every category above: the growing recognition that the decisive edge in modern aerospace increasingly lives in software rather than hardware.
The clearest evidence is the British Army’s ASGARD system, a “digital targeting web” that has compressed corps-level military planning cycles from 72 hours to a single hour and allowed a single corps headquarters to prosecute roughly ten times as many targets in a day, according to Chief of the General Staff General Sir Roly Walker. ASGARD is not a new missile or a new radar. It is an integrated software architecture, built from components including Anduril’s Lattice, Helsing’s Altra targeting software, and a networked command-and-control layer, that fuses existing sensors and weapons into a single, AI-curated decision loop. The equivalent American effort, CJADC2 (Combined Joint All-Domain Command and Control), demonstrated at scale during the U.S. Army’s Project Convergence Capstone 5, pursues the identical goal: a common operating picture that lets any sensor cue any available weapon across every domain and every allied service, with the software doing the correlation work that used to consume days of staff officer time.
The same software-first logic shows up at the individual weapon-system level. Lockheed Martin’s Sanctum-Grizzly-JAGM counter-drone kill chain, assembled from an AI battle manager, an existing radar, an existing missile, and a containerised launcher, went from concept to live-fire intercept in under 45 days, not because any single hardware component was new, but because the software layer connecting them could be built and validated far faster than a from-scratch weapon programme ever could. Northrop Grumman’s newly unveiled Raid Hunter gun-based air defence system follows the identical template: a genuinely mature chain gun already in Army production, paired with new guided ammunition and a networked battle-management controller designed to slot into the wider Integrated Battle Command System architecture.
What is emerging across all these programmes is a genuinely different model of aerospace development. Where the twentieth-century paradigm optimised a single exquisite platform over a decade-long development cycle, the current model increasingly treats hardware as a relatively stable, procurable commodity and concentrates the fastest-moving innovation in the software that ties hardware together, updates every eight to twelve weeks in ASGARD’s case, and improves without ever requiring a new airframe.
5. The Space Layer Becomes Load-Bearing
The fifth force is less visible than fighter jets or drone swarms, but arguably just as consequential: aerospace’s centre of gravity is quietly extending into orbit, and space-based infrastructure is becoming a load-bearing part of systems that have nothing to do with satellites on the surface.
The clearest example is the Proliferated Warfighter Space Architecture (PWSA), the Space Development Agency’s programme to build a constellation of several hundred satellites in low Earth orbit, organised into a Transport Layer that acts as a resilient, low-latency mesh communications backbone and a Tracking Layer purpose-built to detect and follow hypersonic glide vehicles and ballistic missiles from space, closing a genuine gap in existing ground-based radar coverage. The Pentagon has committed roughly $11 billion to PWSA since 2020 and plans to spend close to $35 billion through fiscal year 2029, and the pace of contracting has accelerated sharply through 2026: the Space Development Agency awarded $3.5 billion in December 2025 to Lockheed Martin, L3Harris, Northrop Grumman, and Rocket Lab for 72 additional Tracking Layer satellites, followed by a further $1.75 billion for 36 more satellites in July 2026, explicitly justified as a step toward fielding a functional homeland missile defence layer before the end of the decade.
The reason this matters well beyond the space community is straightforward: a hypersonic glide vehicle’s entire tactical advantage rests on its ability to manoeuvre unpredictably below the coverage envelope of ground-based radar. A space-based sensor layer that can track that vehicle continuously from above closes exactly the detection gap that has made hypersonic weapons so difficult to counter, feeding data directly into the same kind of AI-curated kill chains that ASGARD and CJADC2 are built to manage. In other words, the AI battle management revolution described above and the hypersonic propulsion race described earlier cannot fully succeed without the orbital sensor layer connecting them. Aerospace’s “future,” in the most literal sense, is no longer confined to the atmosphere.
Progress has not been without friction. As of March 2026, a portion of PWSA’s Tranche 1 satellites, launched in September and October 2025, had still not completed on-orbit checkout, a reminder that proliferated small-satellite architectures still face real engineering and schedule risk even as they scale toward hundreds of spacecraft.
6. The Economics of Mass
The final force, and the one that arguably ties the other five together, is economic rather than technical: a growing, industry-wide recognition that the defining tactical problem of the current decade is not whether a given interceptor, sensor, or aircraft works, but what happens when a defender has to engage the fortieth cheap drone of a raid using a fortieth expensive interceptor.
That cost-exchange problem is why guided-ammunition gun systems like Raid Hunter, containerised missile launchers like Grizzly, and directed-energy weapons whose only meaningful marginal cost is electricity, from Israel’s operational Iron Beam to Thales’ RapidDestroyer, which downed 80 drones in a single trial, are proliferating simultaneously rather than converging on one dominant answer. It is also why advanced manufacturing has become a genuine strategic priority rather than a curiosity. Future Military Technologies covered the U.S. military’s growing reliance on 3D printing across the Army, Navy, and Air Force earlier this year, and more recently, the remarkable case of a U.S. Air Force C-17 grounded for seventeen months because no commercial vendor would build a single replacement nose panel for an aircraft whose production line closed in 2015, until the Air Force’s Rapid Sustainment Office used a robotic Incremental Sheet Forming process to manufacture the part in-house.
That C-17 story is worth dwelling on, because it is the quiet, unglamorous flip side of everything else in this article. Every exquisite platform discussed above, the F-47, GCAP’s Tempest, PWSA’s satellites, will eventually face the same arithmetic: a production line that closes, a supplier that walks away from a one-off order, a fleet that has to keep flying decades after the factory that built it goes cold. The same manufacturing ingenuity currently being applied to a 1990s-era transport aircraft will, within a few decades, almost certainly be applied to today’s sixth-generation fighters. Advanced manufacturing is not a side story to the future of aerospace technology. It is the mechanism that determines whether any of the other five forces described here remain flyable long enough to matter.
Where This Leaves the Next Decade
Pulling these six forces together produces a reasonably clear, if unevenly paced, picture of where aerospace technology is heading through the early 2030s.
Crewed sixth-generation fighters will reach first flight, the F-47 in 2028, GCAP’s demonstrator in 2027, before their intended adaptive-cycle engines are ready, meaning early flight testing across multiple programmes will rely on interim propulsion arrangements rather than the definitive powerplant. Uncrewed formations will scale dramatically in this same window, with the U.S. Air Force’s roughly 1,000-aircraft Collaborative Combat Aircraft ambition and the broader rotary and ground-based autonomy programmes discussed above expanding from single-digit test fleets to genuine formations. The space-based sensing layer needed to track hypersonic threats and feed AI-curated kill chains will mature in roughly the same period, assuming PWSA’s Tranche 2 and Tranche 3 constellations clear their on-orbit checkout processes on schedule, itself an open question given Tranche 1’s delays. And the economics-of-mass problem, cheap drones against expensive interceptors, will almost certainly intensify before it resolves, since none of the current answers, guided guns, containerised missiles, or directed energy, has yet been proven at the scale a full peer conflict would demand.
The single biggest constraint across all six forces is not physics. It is industrial capacity and workforce, the same limiting factor that grounded a C-17 for a year and a half over a single sheet-metal panel. Every programme described in this article, from hypersonic glide bodies to adaptive-cycle jet engines to proliferated satellite constellations, ultimately depends on a defence industrial base that has to simultaneously sustain Cold War-era fleets, ramp up next-generation production, and absorb entirely new categories of software-defined, rapidly iterating systems that did not exist in the acquisition playbooks written for previous generations of aircraft.
The future of aerospace technology is not a single breakthrough
The future of aerospace technology is not a single breakthrough waiting to happen. It is six things happening at once, at different speeds, in different domains, occasionally colliding in useful ways, an AI battle manager cueing a hypersonic interceptor whose targeting data came from a satellite that did not exist five years ago, launched from a containerised system built in six months rather than a decade.
What makes this moment genuinely different from previous eras of aerospace innovation is the compression of timelines across categories that used to move independently. Propulsion, autonomy, software, space infrastructure, weapons economics, and manufacturing are no longer separate engineering disciplines advancing on separate schedules. They are increasingly one interconnected system, and the militaries and companies that understand that interdependence, rather than optimising any single piece of it in isolation, are the ones setting the pace for everyone else. The next decade of aerospace technology will belong less to whoever builds the single fastest aircraft, and more to whoever builds the system that connects the fastest aircraft to everything else flying, floating, rolling, and orbiting around it.
