Johan Segertoft, head of the Gripen E program at the Swedish aerospace company Saab, speaks about the new E version in terms that might seem more suited to computers than to military hardware: He talks about software upgrades and data layers, coding and multicore processors. “Our job is to let the pilot make decisions faster than the other pilot,” Segertoft says, and, over the years, Gripen pilots have out-maneuvered a few other hotshot fighters in international exercises. The JAS 39E Gripen, which first flew last December, is the latest version of a Mach 2 fighter that rolled out in 1987.
To an untrained eye, the Gripen E does not look greatly different from its predecessors which flew 50 years ago. Delta-winged, conical-nosed jet fighters that fly at twice the speed of sound and carry a few tons of bombs and missiles under their wings were around in the Vietnam War.
But an aircraft’s effectiveness is no longer determined mainly by how fast it can fly or how quickly it can turn. Now it’s about how well it can interpret information from many different sources and present the data to pilots in ways they can instantly understand and act upon. And while the technology of how to make an airplane powerful, streamlined, and maneuverable is fairly mature, computer technology continues to improve exponentially. So an aircraft’s effectiveness in battle is determined, in part, by how quickly its electronics can be upgraded. And since the middle of the last century, Sweden has been pursuing effectiveness for one very specific battle.
Sweden is a non-aligned country, but although it has never joined NATO, it has been clear since the cold war who its likely adversary would be.
“Being neutral, just next to this huge country, Russia, meant that we have been very focused on defensive counter-air,” says Colonel Torgny Fälthammar, a former Gripen pilot and head of the Swedish air force’s Gripen program. Sweden must be “able to intercept and take down attacking aircraft in order to protect our country and make sure the rest of the armed forces can mobilize.”
Its single focus has given the manufacturer Saab a proclivity for making strange-looking airplanes. Its first aircraft entered service with the Swedish air force in 1941—a relatively conventional dive-bomber, the Saab 17. But after that, Saab produced a sequence of quirky aircraft. There was a small twin-boom, pusher-propeller fighter, the Saab 21, which was upgraded to jet engines after the war, and looked rather like the de Havilland Vampire. Then in 1955 came the Draken—a big, high-speed, high-altitude interceptor. Next, in 1967, Sweden unveiled the Viggen, a strike fighter designed for short take-off and landing, reflecting Sweden’s expectations that its air bases would rapidly be targeted in any war with the USSR.
These aircraft were both striking and unusual—the Draken with distinctive wide, double-delta wings; the Viggen, with canard foreplanes. And both aircraft were groundbreaking in a variety of ways. “What was really cool,” says Richard Smith, Saab’s deputy head of marketing on the Gripen project, “was that [the Draken] had the first data link.” Ground radar stations transmitted tiny packets of data via radio to the Draken: the pilot would be sent tactical information, such as the speed and altitude of the target, or the direction to it, given as a point on the radar scope. The idea was that, in a war with Russia, unsecured voice radio could be easily intercepted or jammed, while digitally transmitted information was more secure. The Viggen, meanwhile, was the first production fighter with a head-up display. It also expanded upon Draken’s data links to include data-sharing between aircraft as well as from ground stations. Also, “it was one of the first aircraft to have a central computer,” says Smith—although it was so long ago that it was known as a “central calculator.”
Then, in the 1980s, came the Gripen.
From the beginning, the Gripen was designed with its electronics in mind. A relatively small aircraft, it is single-engined and weighs about six tons empty. Its similar-looking but twin-engine contemporaries, the Eurofighter Typhoon and French Dassault Rafale, come in at 11 and 10 tons respectively. It was built so small, says Fälthammar, because its designers expected computers to get small enough to fit in it. “That decision was made in 1982,” he says. “And it was risky, because in 1982 the computer technology meeting the requirements regarding performance and miniaturization did not yet exist.” This stripped-back design also helped keep costs down. “An engine is around $10 million, say, so [by using just one] you’ve immediately reduced the price by a sensible amount,” says John Sneller, head of aviation at the defense analysis company Janes.
A further advantage is that Saab is able to avoid many of the political considerations that tied up the Eurofighter project. That meant it could buy off-the-shelf parts—for instance, it uses a General Electric F404 engine, built under license by Volvo. Using premade parts of known reliability reduces risk and cost. “It’s adding risk on risk if you’re building a new aeroplane and a new engine at the same time,” says Sneller. “You saw it with the F-35, A400M, Tornado, and Typhoon. Trying to do everything to the top level of tech at the same time is a high-risk approach.” But in international collaborations, it can be hard to avoid.
A defense industry worker and former Royal Air Force fighter pilot who asked to remain anonymous describes the pressures that influence international projects. “If you look at Typhoon,” he says, “four countries each make a bit of the radar; the engines are made by Eurojet.” Politics are involved: “ ‘I need jobs for Lancashire, I need jobs for central Italy’—Saab can be a lot leaner and more agile and [they] don’t have to try to keep everyone happy.”
The Gripen E project began in 2010 and—even more than the earlier versions—was built around its electronics suite. “Gripen E has some of the coolest sensors available, but to use them in the best way you need computer power to fuse all that information and provide it to the pilot in such a way that he can complete his task,” says Segertoft.
Processing power has other uses: Modern fighters can identify other aircraft by the particular radar signatures of the fan blades of their engines. “It started with the F-15,” says the former RAF pilot. “The radar blip has fuzzy lines from the engine modulation. If you have enough computing power, you can work out it’s a Russian engine in a MiG-29.”
He continues: “An F-35 is flying around looking for things. It’ll have flight plans for airliners. [The pilot will] get a blip. It will say that’s in an airway. There’s a 92-percent chance it’s an airliner. Whereas if it’s a small blip, I might want to spend a lot more effort finding out if it’s something I should be interested in.”
Modern electronic warfare measures rely heavily on computer power. “One thing that sets the Gripen apart from its direct competitors is the degree to which the entire aircraft and systems were designed around the electronic warfare suite,” says Justin Bronk, a research fellow at the Royal United Services Institute specializing in combat airpower and technology. That includes both electronic countermeasures (ECM) and counter-countermeasures. If an enemy radar is using a certain frequency, your ECM can blast that frequency at it, so that it is deafened with noise; but it can deploy counter-countermeasures, such as switching frequencies, so your ECM needs to be clever at identifying those frequency switches and keeping up.
And each year, as the state of the art improves, the complexity of these systems explodes. “In embedded systems like cars, satellites, aircraft,” says Segertoft, “if you needed 100,000 lines of code a few years ago to solve a problem, it’s a million now, and it’ll be 10 million in a few years.”
Software upgrades are traditionally carried out cautiously. It’s one thing if you install a new version of Excel and your computer freezes in the middle of an important piece of work; it’s another if your $100 million fighter does, especially if it happens just when you need it to help you avoid an incoming missile. But Saab has a different philosophy.
“We needed to think differently because [each software or hardware] upgrade was taking too long to get out,” says Segertoft. The idea behind the Gripen E was to make that process smoother.
Segertoft’s favorite analogy is the iPhone. When you upgrade to a newer version—with more processing power and a better camera—and then you boot it up, he asks, “Will you as a user notice? No, because the software is still working.” You can plug and play new software, secure in the knowledge that it will work on the hardware, and you can install new hardware, knowing it is compatible with the software. Gripen E, he says, is the same: You can put in new hardware, and it’ll run the same software; you can switch software modules, and they’ll cooperate with each other and with the hardware.
The question is how can you—absent the years of trial and error for each specific upgrade—test the whole system and be sure that it is robust in this general way. “That’s the meat in the pie,” agrees Segertoft. “How do you verify, to aviation standards, that it is robust to an infinite amount of configurations? Add a new software, remove it, modify it?” It requires building from the ground up, he says; all the components are, not physically separated, but logically separated—as with the iPhone, the software “apps” can be removed, replaced, or upgraded without affecting the others. “I heard colleagues in the industry trying something similar, and they gave up because they deemed it impossible.” It was “naive and brave” to try to build it, he says—“it’s something inherently Swedish—naive enough because we don’t know what it means but brave enough to try anyway.
“The initial efforts are almost impossible, but when you’ve done it you’ve done it,” adds Segertoft. For example: when the Gripen E was first designed, there were no “multicore processor” computers—that is, computers which had two or more processors acting in parallel, and which could carry out many more operations per second than single-processor machines. But the E’s modular design—the fact that, like an iPhone, its hardware could be upgraded while leaving the software in situ—meant that multicore processors could be added without fuss. Says Segertoft: “We added them, and it wasn’t a big deal, because we had the platform.”
Bronk says that the EW suite in particular seems to be exceptional—outsiders don’t know, exactly, because these things are secret, but “on the rare occasions where Gripens [C variants] have turned up in something like war mode, the Typhoon pilots go ‘Wow, that seems really impressive.’ ”
It is also a formidable aircraft in more traditional respects—agile, fast, capable of cruising at supersonic speeds without afterburner even when carrying external weaponry and fuel tanks. Because of its small size, it has a smaller radar cross-section. “You have to get closer or put out more energy in your own radar to detect a Gripen,” says Fälthammar. That means that the Gripen can get closer before launching its own missiles, giving it “a few more kilometres of effective range.” The former RAF pilot adds that being small also makes it harder to spot visually, and being maneuverable gives it a few extra seconds in which to turn and outrun incoming missiles.
The Gripen has its weaknesses. Because of its small size, it pays a higher cost—in range and speed—than one of its larger rivals for a similar amount of ordnance hung from its hardpoints. For all that the fighters are designed to catch and intercept Tupolev Bears over the Baltic Sea, they can’t actually keep up with them for very long. The Bear can just keep on flying, for hours and hours, and a Gripen will quickly run out of fuel. One engine saves on cost, but means less redundancy and less spare power for other systems, such as radar.
But Gripens have performed well in international operations, such as Red Flag. Said Swedish General Lennart Pettersson in 2004: “Although some aircraft flown in these exercises may have had a slightly better thrust-to-weight ratio than our aircraft, Gripens still managed to get behind the F-16s to make use of both their IR-missiles and guns. The small visual signature and excellent agility of the Gripen proved to be a considerable advantage in a dogfight.”
Perhaps more tellingly, a Chinese People’s Liberation Army Air Force report about a 2015 engagement exercise between PLAAF J-11s, the Chinese version of the Su-27 Flanker, and Thai Gripen Cs found that the J-11s won comfortably in dogfights—their greater thrust-to-weight ratio allowed them to turn faster and get shots off with infrared missiles and cannon. They shot down 16 Gripens with no losses. But in beyond-visual-range engagements, the Gripens were deadly: They shot down 41 Flankers against nine losses.
The Gripen has been an export success because it is extremely cost-effective for countries who want roughly what Sweden wants from its fighter aircraft. “In incredibly blasé terms, Gripen gives you about 85 percent or 90 percent of the highest level abilities, at about 50 percent of the cost,” says Bronk. “If you’re an air force that wants to generate a reasonable number of sorties from a limited budget, and your main concern is air defence and policing, then it’s frankly an obvious choice.” The Gripen C has found customers in Czechia, Hungary, Thailand, South Africa, as well as Sweden. Britain’s Empire Test Pilots’ School has had a Gripen D for training pilots since 1999, because, says Fälthammar, the simple, responsive control system is highly regarded.
Sweden’s neutrality has also been an advantage for some export deals. If you are a non-aligned country yourself—say, Brazil—and you buy the F-35, you are making a geopolitical statement about where your allegiances lie. “Brazil doesn’t want to be too close to the U.S., or the Russians, or the Chinese,” says Sneller. “[Gripen] is therefore attractive to non-aligned countries, like Austria and Finland.”
Previous Saab aircraft have not sold especially well overseas because their individuality counted against them. “If you look at the Viggen, its cockpit had a totally different setup,” says Bronk. “If you were used to any other NATO aircraft, [it] would be completely baffling. You’d have to completely revamp your training.” Gripen, though, is much more manageable for pilots of other modern aircraft and has sold well. The Gripen C is expected to remain in service until the 2030s, says Fälthammar—“it’s robust, it’s a very mature and good fighter aircraft”—but Sweden is taking delivery of the first 60 Es now, and Brazil has ordered 36. There are hopes that Finland and Canada may be future customers.
So far, says Fälthammar, the E has not flown much against other aircraft since it is still under development. But he waxes lyrical about how the aircraft is to fly: “It’s such a lovely experience. When you go down into the cockpit, it’s high-tech, it’s sleek, it’s like sitting down in a modern sports car.”
Except, unlike most new cars, the upgradable design means that countries won’t have to trade a Gripen in for a new model every few years—a deal that’s tough to beat.
Tom Chivers is an award-winning science writer and science editor at UnHerd.com.