In a close-in fight the old rule was brutally simple: to shoot at something, you had to point the aeroplane at it. The gunsight, and later the head-up display, sat fixed on the centreline, and everything a fighter pilot could kill had to be dragged into that narrow box of glass. Whole generations of air combat manoeuvring existed for no other reason than to get the nose onto the target.

The helmet-mounted display broke that rule. Put the display on the pilot's head instead of on the canopy rail, track where the head is pointing, and the missile seeker can be slaved to the pilot's line of sight. Look at the target over your shoulder, press a button, and the weapon looks there too. The aeroplane never has to turn.

It sounds obvious. It took roughly forty years to get from the first crude experiments to a helmet that could reliably do it under 9g, and the story runs through a South African missile programme, a Soviet fighter that badly frightened NATO, and finally an American fighter that deleted the head-up display altogether.

Before the helmet, there was the glass

The head-up display solved a real problem. Instead of forcing the pilot to look down at instruments during the few seconds that matter, it collimated airspeed, altitude, heading and weapon symbology at optical infinity on a combiner glass in front of the windscreen. Look out, fight, and the numbers float on the world rather than in a dial.

What the HUD could never solve was geometry. It is bolted to the aircraft, so its picture is the aircraft's picture. A target at 60 degrees off the nose simply is not in it. Early infrared missiles had narrow seeker fields of view and had to be told roughly where to look, which meant the aircraft had to be pointed roughly at the target before the shot was even possible.

The idea of moving the display onto the head is older than most people assume. In 1962 Hughes Aircraft showed the Electrocular, a compact cathode-ray-tube monocular display that reflected a television signal onto a transparent eyepiece. By the early 1970s the US Navy was flying Honeywell's Visual Target Acquisition System on the F-4J, and VTAS went through the ACEVAL/AIMVAL trials on F-14s and F-15s between 1974 and 1978, where it was praised for exactly the thing it promised: cueing off-boresight missiles. The United States then did not field it widely, beyond integration into late-model Navy Phantoms carrying the AIM-9 from 1969.

South Africa built it, the Soviets weaponised it

While the Americans were shelving VTAS, the South African Air Force was quietly putting the concept into squadron service. From 1975 its Mirage IIICZ and Mirage F1AZ fighters carried a locally developed helmet-mounted sight integrated with the Armscor V3A heat-seeking missile. The point was never elegance. It was that a pilot could take an off-boresight shot without first manoeuvring into the classic firing cone behind the target, and the system saw combat over Angola.

The Soviet Union arrived at the same conclusion and built the most consequential version of it. The Shchel-3UM helmet sight, a 1981 design, was fitted to the ZSh-5 and later ZSh-7 helmets and paired with the R-73 missile, known to NATO as the AA-11 Archer. The MiG-29 entered service in 1985 with that combination, and so did the Su-27. In a turning fight it meant a Fulcrum pilot could look sideways, uncage a very agile missile, and shoot at an angle no Western fighter could answer.

NATO found out how serious this was in the most direct way possible. After German reunification, the unified German Air Force inherited and operated ex-East German MiG-29s, and Western air forces got to fly against the helmet and the Archer instead of guessing about them. Several nations launched programmes specifically to counter the combination.

JHMCS: the West catches up

Israel got there first in the West. Elbit began developing the DASH in the mid-1980s against an Israeli Air Force requirement for the F-15 and F-16, with the first design in production around 1986 and the GEN III helmet arriving in the early-to-mid 1990s. DASH built the optics and the position-sensing coils into the helmet shell itself, using a spherical visor to give the pilot a collimated image, and talked to the weapon system over a MIL-STD-1553B bus. It became the baseline technology for the American programme that followed.

The United States had been pursuing a helmet paired with ASRAAM alongside the UK and Germany. Technical difficulties led the Americans to walk away from ASRAAM and instead fund two things in 1990: the AIM-9X, and the Joint Helmet Mounted Cueing System. JHMCS came out of Vision Systems International, a joint venture between Rockwell Collins and Elbit, drawing on both DASH III and the Kaiser Agile Eye.

It reached squadrons in November 2003, with the 12th and 19th Fighter Squadrons at Elmendorf AFB in Alaska. The Navy had used the F/A-18C as its test platform but fielded JHMCS first on the Super Hornet, also in 2003. Unlike DASH, JHMCS bolts onto modified HGU-55/P, HGU-56/P or HGU-68/P helmets rather than being built in, and the design is around 95 percent common across every aircraft that carries it.

What Godfrey is describing is the transition itself. For a decade or so the helmet did not replace the HUD, it supplemented it, and the two displays divided the work: flight-critical data stayed on the glass, the aiming moved to the head.

What off-boresight actually buys you

Pair JHMCS with the AIM-9X, a short-range weapon with a focal-plane-array seeker and a thrust-vectoring tail, and a pilot can designate a target up to 80 degrees either side of the nose. That is not a marginal improvement on a fixed gunsight. It is most of the forward hemisphere, and it turns a lot of previously survivable geometry into a shot.

The extreme demonstration came in March 2009, when a Royal Australian Air Force F/A-18 using JHMCS fired an ASRAAM at a target located behind the wing-line of the shooting aircraft, with the missile locking on after launch. The weapon was fired essentially over the shoulder.

There is a cost to this, and it is not paid in hardware. High off-boresight shooting makes the merge far more lethal for both aircraft, which is one reason modern fighter training puts so much emphasis on never getting there in the first place. The helmet did not make the dogfight safer. It made it shorter.

The F-35 takes the HUD out entirely

Vision Systems International, working with Helmet Integrated Systems, built the Helmet Mounted Display System for the F-35, and Lockheed Martin took a decision no other modern fighter programme has taken: there is no head-up display in the cockpit. Everything the HUD used to show is drawn on the visor. That makes the F-35 the first tactical fighter in 50 years to fly without one. The system had a famously difficult development, to the point where a BAE Systems alternative was lined up, but the problems were worked through and HMDS was declared ready for delivery in July 2014.

The helmet also gets the picture from the AN/AAQ-37 Distributed Aperture System, six infrared sensors spaced around the airframe that together give unobstructed spherical coverage with no aiming or pilot input required. DAS feeds missile detection and tracking, aircraft detection and tracking, and imagery for the cockpit displays and for night vision on the helmet. This is the origin of the much-repeated line about F-35 pilots looking through the floor of the aircraft: look down, and the helmet shows you the DAS image of what is underneath.

Not everyone is convinced the trade was free. The pilots who fly it are, as usual, the most useful witnesses.

The helmet is also the reason the F-35’s sensor suite is worth anything to the person flying it. A spherical infrared picture is useless if the pilot cannot get at it quickly, and the visor is the fastest display surface there is.

Why the hard part is the head, not the optics

The physics of a helmet display are unforgiving in a way the marketing rarely admits. The helmet position is what points the missile, so it has to be calibrated and sit securely on the pilot's head. Head anthropometry and facial anatomy make individual fitting a crucial factor, and helmet shift or misalignment produces an inaccurate picture. Precision is measured as the angular error between where the pilot is actually looking and the cue the system derives, and latency is how far behind the head the cue lags.

Then there is weight. Total helmet mass and its centre of gravity are the largest problem fighter HMD designers face, because everything on the head is multiplied under g and has to stay survivable through an ejection. This is precisely why elaborate helmet displays appeared on helicopters first: the US Army fielded the AH-64 Apache in 1985 with the Integrated Helmet and Display Sighting System, a Honeywell monocular with a 40-by-30-degree field of view slaved to a nose-mounted thermal camera. A helicopter cockpit does not pull 9g.

Tracking where the head is pointing has been done five ways: inertial, optical, electromagnetic, sonic, and hybrid combinations. The MiG-29 and its Archer used an optical system with infrared emitters, which is sensitive to sunlight and other heat sources. JHMCS kept DASH-style electromagnetic sensing, which requires precise magnetic mapping of the cockpit to account for ferrous and conductive material in the seat, sills and canopy. Newer systems such as the Thales Scorpion use a hybrid optical-inertial tracker to get low latency and high accuracy together.

The current field

Scorpion is the interesting outlier. Introduced by Gentex and Raytheon in 2008, it won the USAF Helmet Mounted Integrated Targeting competition in 2010 and was deployed on the A-10 and F-16 in 2012, with the helmet-display business later acquired by Thales. It was the first HMD deployed that could display full-colour conformal symbology, and crucially it was designed to clip onto standard-issue HGU-55/P and HGU-68/P helmets with no special fitting, using a light-guide optical element and software correction to compensate for wherever the display happens to sit. The USAF and Air National Guard have been replacing JHMCS with Scorpion on the F-16C.

Elsewhere: the Eurofighter Typhoon uses BAE Systems' Striker and the later Striker II, capable of both raster imagery and cursive symbology with provision for embedded night-vision goggles. The Gripen C/D flies the Cobra, a refinement of Striker developed by BAE with Denel Cumulus. France paired the Topsight, and later TopOwl derivatives, with the thrust-vectoring MICA on the Rafale and late Mirage 2000s. Elbit's Targo II serves Rafale F3R customers including Qatar and India. DASH IV is integrated on India's HAL Tejas.

The technology has also escaped the fighter world entirely. The first civil application was Elbit's SkyLens wearable HUD on the ATR 72 and ATR 42, where the point is not shooting anything but flying an approach in poor visibility with the symbology in front of your eyes rather than on a glass panel you have to look through.

Where it goes next

Two things are still on the list. Eye tracking, which would measure the point of gaze relative to the direction of the head, is not used in aircraft today: current helmets know where your head is pointing, not where your eyes are looking, so they place a predicted impact marker on the line between eye and target and trust you to keep your eye aligned with the sight. And direct retinal projection, painting the image onto the retina with a low-powered laser, remains experimental.

Neither changes the basic bargain. A helmet display buys you the ability to aim at things the aeroplane is not pointing at, in exchange for carrying an expensive, precisely fitted, carefully balanced piece of optics on your head through every sortie of your career. Fifty years of fighter pilots have decided that is a trade worth making.

Sources: Wikipedia (Helmet-mounted display; AN/AAQ-37 Distributed Aperture System), Hush-Kit pilot interviews, Rockwell Collins and Vision Systems International material cited therein.