When Good Weapons Development Goes Bad
Dan Ward recalls the best programs’ worst failures
U.S. Air Force Lt. Col. Dan Ward is a huge advocate of faster, cheaper and more disciplined weapons-development. Forget spending 20 years and half a trillion dollars inventing a big, overly-complex stealth fighter that might not even work when it finally enters service … late.
Ward argues that the Pentagon gets better results when it keeps its goals simple, its deadlines firm and its engineering teams and their budgets small. “The most successful project leaders from government and industry alike tend to deliver top-shelf stuff with a skeleton crew, a shoestring budget and a cannonball schedule,” he writes in his new book F.I.R.E: How Fast, Inexpensive, Restrained and Elegant Methods Ignite Innovation.
But sometimes even small, disciplined teams fail. “Developing new technology programs is a messy business,” Ward explains as he details two military tech efforts that went wrong despite the best intentions.
The great little fighter no one wanted
In the late 1970s, U.S. jet-maker Northrop designed the F-20 Tigershark fighter quickly and cheaply, resulting in an airplane that was also, well, quick and cheap—and seemingly perfect for small air forces with modest means.
The F-20 wasn’t the fastest or longest-range fighter in the world nor the most heavily-armed. But it was elegant, reliable and easy to fly.
Northrop reasonably expected to sell hundreds of the little fighters—especially since Pres. Jimmy Carter’s administration banned the export of the fully-equipped F-16, Lockheed’s bigger, more complex and more expensive plane.
But then Ronald Reagan took office and ended the F-16 ban. That doomed the F-20. “Nobody in the entire world bought one,” Ward recalls. All those air forces wanted F-16s, instead—even if they were pricier and harder to use.
“It is not enough for [a new technology] to be affordable, available and reliable,” Ward explains. “It is not enough for it to be easy to use and maintain. It’s even not enough for it to be demonstrably well-suited to the customer’s needs. If we produce something that is fast, inexpensive, restrained and elegant, we run the risk of being completely rejected by customers precisely because they would rather have a more expensive, complicated product.”
“It doesn’t matter if they are making a good choice or not,” Ward writes. “It’s their choice.” And in rejecting the F-20, air arms did make a mistake, Ward asserts. “If this little jet had won, it would have won big.”
Frankenstein’s air-defense gun
In the early 1970s, the U.S. Army was determined to develop a new anti-aircraft gun to defend tank columns—and to do it quickly and cheaply.
The Sergeant York air-defense gun “was to be primarily assembled out of existing, mature components with a proven track record and known functionality,” Ward recalls. “This approach was intended to save time and money while also minimizing complexity.”
But man oh man did that strategy backfire. The Army chose to use the chassis of the 1950s-vintage M-48 tank, rated for a total vehicle weight of just 50 tons. But the Sergeant York weighed 60 tons, making it very very slow.
Worse, the ground combat branch borrowed the F-16’s tracking radar to guide the gun. “Just as the M-48 chassis was a poor fit for the [Sergeant York’s] actual needs, the jet fighter’s radar proved to be ill-suited to the mission.” It was best at following fast-moving targets while fitted to an equally fast plane. But the Sergeant York moved slow, and many of its targets were also fairly sluggish—helicopters, for instance.
In one notorious test that Ward describes, an Sergeant York prototype mistakenly locked onto a portable toilet, “apparently misidentifying the latrine’s rotating fan as a low-flying helicopter.”
“This was considered an undesirable feature,” Ward writes.
The bottom line was that in their zeal to stay on schedule, the gun’s developers blithely ignored performance problems arising from mismatched components. “Right idea, wrong execution,” Ward quips.
By the time the Army canceled the Sergeant York in 1985, the program had spent nearly $2 billion. But the military at least had the good sense to end the effort rather than try to fix it.
These two failures aren’t arguments against self-discipline in weapons development. Fast, inexpensive, restrained and elegant is still the way to go, Ward argues—and even the failures resulting from this approach can be useful. “An epic fail is one that costs a lot and teaches little. An optimal fail is the reverse—it teaches a lot and costs a little.”