F-35 LIGHTNING II

The F-35 carries the AN/APG-81 AESA radar, the AN/AAQ-37 Distributed Aperture System (DAS), the AN/AAQ-40 Electro-Optical Targeting System (EOTS), and the AN/ASQ-239 electronic warfare suite. Individual coverage of these sensors is already abundant. What's missing from most writing on the subject is how they relate to each other.

On a traditional fourth-generation fighter, the radar has its own display and controls, the infrared search and track has its own display and controls, and the electronic warfare warning system has its own display and controls. The pilot stares at all of them and builds a mental picture: what's out there, where, and whose side it's on. Under high-G maneuvering with threats appearing from multiple directions, this cognitive workload grinds against the ceiling of what human brains can handle. The F-35 eliminated this burden at the architectural level. Its fusion engine pulls raw data simultaneously from more than five onboard sensors and two offboard data links, runs it through correlation, weighting, multi-spectral triangulation, and evidence-based reasoning algorithms, and compresses everything into one single situational awareness picture. On the helmet display the pilot sees something like "enemy aircraft 82 nautical miles ahead, heading 270, speed Mach 1.2, threat level medium, weapons envelope marked." Not separate fragments from separate sensors. One picture.

Processing priority gets assigned automatically based on distance. Close threats get more sensor resources and more computing cycles. Distant targets get deprioritized. What this amounts to in OODA-loop terms is that the fusion engine handles the Observe and Orient steps, the two steps that eat the most time, freeing the pilot to focus on Decide and Act.

In early software versions, this same fusion system kept merging two physically separate targets into one track. The problem had a name: Track Merge. In a ground attack scenario, two vehicles two hundred meters apart getting fused into a single target meant the weapon went to the wrong spot. Later software updates corrected the separation algorithms, and Track Merge incidents dropped. What stuck around is the engineering tension underneath: as automation deepens, recovering from automation errors gets harder. And the entire design trajectory of the F-35 keeps pushing toward more automation, not less.

There's another dimension here that rarely gets discussed. The fusion system also ingests offboard tracks and measurements arriving over Link 16 and MADL. A single F-35's battlefield picture is never just the output of its own sensors. It's the collective product of sensor data from the entire formation, weighted and fused together. When multiple aircraft each detect the same target from different angles, the fusion engine triangulates an exact three-dimensional position automatically. Even if one F-35's sensor only gets bearing without range, a wingman detecting the same target from another angle gives the system enough to solve for precise coordinates. A whole formation can achieve accurate targeting without anyone turning on a radar. Zero electromagnetic radiation. The opposing force's radar warning receivers hear nothing. A group of jets that are electromagnetically invisible, achieving situational awareness accuracy exceeding any fourth-gen fighter running full-power radar.

No HUD on the F-35. That transparent glass combiner in front of the cockpit, the thing that's been standard equipment on fighters since the 1960s, is gone. Everything goes onto the helmet visor: flight data, weapons information, fused sensor imagery. Vision Systems International, an Elbit Systems subsidiary out of Israel, makes it. Over $400,000 per unit.

A HUD shows information in only one direction: straight ahead. The helmet shows it everywhere the pilot looks. Combined with DAS, which provides 360-degree infrared coverage, the pilot can look down between their knees and see the ground through the airframe in infrared. Eliminating the HUD was a gamble when the decision was made.

And the Gen III helmet took years of misery to get working. Early versions had latency: the DAS infrared video feed reaching the visor lagged behind the pilot's head movement, creating a nauseating "jelly effect" when turning quickly. Night flights produced a stray light artifact pilots called "green glow" on the visor, bad enough to compromise reading the outside environment. Weight and center-of-gravity were a chronic headache too. In an ejection, the heavy helmet would subject the pilot's neck to forces beyond the safe limit. Lockheed Martin and Elbit ended up adding weight-reduction padding and rebalancing the CG inside the helmet shell. Current production helmets have mostly resolved these issues. $400,000 each, and it took close to a decade to sort out.

Before every mission, an F-35 loads a pre-programmed data package called a Mission Data File, or MDF. Inside it: the electromagnetic signal signature library for every known threat system in the operating area. Enemy radar frequencies, pulse repetition frequencies, scan patterns, missile guidance signal waveforms, plus friendly platform identification parameters.

If the MDF doesn't contain the signature for a particular air defense radar model, then even though the F-35's sensors physically detect that radar's signals, the fusion engine can't automatically classify it as a known threat or match the right countermeasure response. The pilot gets an icon labeled "unknown RF emitter" instead of "SA-21 missile system, 400-kilometer range, search mode." On a battlefield that gap can kill you.

Each F-35 customer nation gets its own tailored MDF, compiled and updated by a U.S.-led reprogramming laboratory. How quickly that MDF gets updated determines whether the F-35 can effectively counter new threats. Adversary fields a new air defense system, MDF hasn't caught up yet, the F-35's recognition efficiency against that threat drops hard.

And then there's the other layer. Which threat data gets included in the MDF shipped to a given country, which gets left out: that's Washington's call. The combat capability ceiling of every F-35 user nation is set by what's inside its MDF. Israel understood this from the start and built an independent mission system and partial MDF self-compilation capability into the F-35I "Adir." Among all F-35 operators worldwide, Israel is the only one with that kind of autonomy. For everyone else, the MDF they receive is the combat capability they get.

The Multifunction Advanced Data Link, MADL, is a fast-switching narrow-beam directional data link built for stealth aircraft. Ku band. Phased array antenna assemblies transmit tightly focused signals. Where Link 16 broadcasts in every direction, MADL works like a very thin flashlight beam pointed at a specific receiver. Enemy electronic intelligence systems have an extremely hard time picking it up or jamming it.

Spread-spectrum modulation, precision beamforming, fast frequency hopping. A four-ship of F-35s can share all raw sensor data at high speed while looking, to the outside world, like they're maintaining total radio silence. Up to eight aircraft can form a mesh network within line of sight. One aircraft detects and cues. Another attacks. A third handles electronic warfare suppression.

That kill chain crosses from the air domain to the surface domain, and the F-35's role in it resembles a scout and a battle manager more than a striker.

MADL has a structural limitation. It basically only works between F-35s. The B-2 is compatible. The F-22 was supposed to get MADL, but that upgrade was canceled around 2010 on cost grounds. So the F-22 is stuck with its own IFDL data link, which only talks to other F-22s. The F-22 can't even transmit on Link 16. America's two fifth-generation fighters cannot conduct native stealth communication with each other. Northrop Grumman's Freedom 550 software-defined radio can translate between IFDL, MADL, and Link 16, at the cost of added complexity and added latency. The decision to cancel the F-22 MADL upgrade saved money at the time. The consequences of that decision are still causing problems in U.S. joint operations planning in 2026.

The avionics on the F-35 draw enormous power. All sensors, the fusion engine, comms, and the EW suite running at full tilt generate far more heat than anything on a traditional fighter.

The engineering solution: use the onboard fuel as a heat sink. Avionics heat gets transferred to the fuel through liquid cooling loops. The fuel absorbs the excess thermal energy. Then the heated fuel goes into the engine and gets burned. Elegant in concept. Counterintuitive in consequence. As the mission goes on and fuel gets consumed, the available heat sink shrinks. Late in a mission, with fuel running low, the avionics system's ability to run at full power becomes thermally constrained. The moments when the pilot most needs every sensor online, returning to base, fuel tight, possibly under pursuit, are exactly the moments when thermal margin is thinnest.

The F-35A, F-35B, and F-35C share roughly 80% of their parts.

The B variant's lift fan eats up a huge volume of space and structural margin behind the cockpit. Maintaining structural commonality across all three variants meant the A and C had to work around that space too. The A could have had a bigger internal weapons bay, or carried more fuel internally, if it didn't need to share its basic airframe geometry with the B. The C needed a bigger wing and beefier landing gear to handle catapult launches and arrested landings on a carrier, and that extra weight constrains its top speed and sustained-G performance by measurable amounts.

Tri-service joint development was supposed to save money through scale. Cramming Air Force high-speed intercept requirements, Marine Corps vertical landing requirements, and Navy carrier operation requirements into the same airframe generated engineering complexity that, at the aggregate level, pushed development cost and schedule in the opposite direction.

Had the U.S. developed three separate fifth-generation fighters for three separate services, total R&D spending would almost certainly have been higher than the F-35 program. The joint approach means every single aircraft absorbed compromises. The payoff is one R&D pipeline, one training system, one global logistics network instead of three of each. Whether the math works out is still unsettled. It's a question nobody can run a controlled experiment on, so it probably stays unsettled.

The Block 4 upgrade package covers the new AN/APG-85 radar, an improved electronic warfare suite, a next-generation DAS, and integration capacity for a large number of new munitions. In 2023 the Government Accountability Office reported that Block 4's estimated cost had swollen from $10.6 billion to $16.5 billion, with the completion date slipping from 2026 to 2029 at the earliest. By September 2025 came a further announcement: Block 4 would be cut in scope and delayed again. The TR-3 hardware refresh required to underpin Block 4 ran into severe regression testing issues all on its own, halting aircraft deliveries from July 2023 to July 2024. For a full year, the largest fighter production line on the planet sat in delivery freeze.

Reports surfaced in 2025 that new-production F-35s were being delivered to the U.S. military without radars. The APG-85 was still in development. The old APG-81 production line was already in transition toward the new model. The Air Force confirmed that Lot 17 jets were going out with APG-81s installed, and then said that further modernization details were classified.

At the root of Block 4's delays: the F-35 mission system software base has reached millions of lines of code. At that scale, integrating each new feature increment demands massive regression testing to make sure the new code hasn't broken anything in the existing functionality. TR-3 brought in a completely new Integrated Core Processor with far more computing power than the previous generation, and that meant every existing software module needed full regression verification on the new hardware. The project planners badly underestimated how much work that verification would take. Writing the new functionality might take three months. Proving it didn't damage what was already there can take two years. That is where the Block 4 schedule has been bleeding out, not in any one subsystem's technical barrier.

From day one the F-35 shipped with a comprehensive logistics management platform called ALIS, the Autonomic Logistics Information System. It was supposed to handle fault prediction, spare parts management, maintenance scheduling, mission planning, and flight data recording, all integrated.

What it actually did: run slowly enough that maintainers went back to pen and paper. Crash regularly. Route spare parts requests to the wrong locations because of database errors. Take so long to boot up its mission planning module that it affected sortie generation. And underneath all of that was an architecture-level problem. ALIS was centralized. Maintenance terminals at the flight line had to connect to a central server over the network to complete many basic tasks. Forward-deployed, where network connectivity is unreliable, that dependency became an operational bottleneck.

The Pentagon eventually decided to scrap ALIS entirely and replace it with a new system called ODIN, the Operational Data Integrated Network, with rollout starting in 2020. ODIN uses a modular, decentralized architecture that lets more operations happen locally and offline. Migrating a logistics management system already deployed to over a thousand aircraft in more than a dozen countries is itself a massive engineering undertaking.

How many jets a squadron can put in the air on a given day depends not just on whether the aircraft work, but on whether the digital logistics ecosystem around them works. If the logistics software can't get the right part to the right place on time, the jet sits on the ramp waiting. The ALIS experience in the F-35 program left a deeper mark than any test flight incident. The United States spent hundreds of billions building the most advanced fighter on earth and then watched a bad logistics application drag sortie rates through the mud.

The Israeli Air Force first took the F-35 into combat in 2018, striking targets in Syria. After that, U.S. forces employed the type in Afghanistan, Iraq, Yemen, and the 2025 military operations against Iran. During Operation Midnight Hammer in 2025, F-35As and F-22s launched from regional land bases to lure and exhaust Iranian surface-to-air missile fire, opening a corridor for B-2 stealth bombers to penetrate. The Israeli F-35I "Adir" shot down an Iranian Yak-130 in air-to-air combat, the first time an F-35 downed a manned aircraft. In March 2026, a British F-35B shot down a hostile drone over Jordan. In January 2026, F-35s participated in U.S. operations against Venezuela, tasked with disabling air defenses to clear a path for helicopters.

These engagements cluster in three areas: suppression/destruction of enemy air defenses (SEAD/DEAD), situational awareness distribution in high-threat environments, and networked command-node functions within the joint strike chain. Traditional dogfighting barely features in the F-35's combat record.

In 2025, the F-35 accumulated nearly 5,000 mishap-free flight hours during a single U.S. Navy deployment. For a type whose reliability has been questioned for years, that figure shows maturity climbing.

Turkey got kicked out of the F-35 program for buying the Russian S-400. Turkey was not only a customer but a supply-chain participant, manufacturing parts for the airframe. Lockheed Martin had to scramble for replacement suppliers on multiple components and subsystems.

Entering the F-35 ecosystem means strategic alignment with the United States. Leaving means losing access to an entire generation of air combat capability. Israel's MDF autonomy, Turkey's expulsion, Switzerland cutting its buy over budget pressure: put these events side by side and the F-35's function as a geopolitical lever becomes obvious. A weapons program that simultaneously serves as an alliance management tool and a technology access gate is an extremely rare thing in post-Cold War arms sales. After getting ejected, Turkey pivoted to its indigenous fifth-generation fighter, the TF-X (since renamed KAAN). That program's current progress and technical maturity still have considerable distance to cover compared with the F-35.

Lockheed Martin delivered a record 191 F-35s in 2025, breaking the previous record of 142. Global fleet approaching 1,300 aircraft across 12 nations. Northrop Grumman has shipped its 1,500th F-35 center fuselage from the automated assembly line at Palmdale, which rolls one out every 30 hours. AR/VR tooling cut assembly time by 35%. Lots 18 and 19, signed at roughly $24 billion for up to 296 jets, represent the biggest production contract in the program's history. Denmark added 16 more in October 2025, bringing its total to 43. Canada confirmed 88 in 2023.

Twelve countries' training systems, logistics chains, data standards, operational doctrines, and mission planning software are now deeply embedded in the F-35 ecosystem. Simulator data packages, mission planning interface protocols, weapons integration standards, allied encrypted data-sharing architectures: all built around the F-35. Switching costs are so high that switching is effectively not an option. Meanwhile in 2025 the Pentagon cut domestic F-35 orders in the FY2026 budget from 74 to 47, and Congress left that cut in place. At the same time the $24 billion Lot 18/19 international contract was the largest ever. Domestic procurement shrinking while international procurement expands. Both curves at the same time.

Total lifecycle cost from the program's 1990s start through projected retirement in the 2070s is currently estimated at about $2.1 trillion. Roughly half of that figure is inflation.

Where the F-35 Lightning II stands in 2026 is clear enough. Sensor fusion changed how air combat works. Twelve nations are embedded in the ecosystem. 1,300 aircraft in service make any alternative economically unviable. Block 4's full promise has been dragging for a decade without delivery. The logistics system went through a complete do-over. Every A model with engine costs over $100 million. All of this is public, known, audited repeatedly.

The aircraft keeps selling. 191 deliveries in 2025 broke the record. Denmark added. Canada confirmed. Lots 18 and 19 produced the largest production contract in program history. Defense ministries voted with their budgets.