HEADLINE: FAILURE INC.; DISASTER IS A CERTAINTY, AND ONE LITTLE COMPANY THRIVES ON IT
BYLINE: By RUSTY WESTON, San Francisco-based Rusty Weston writes about
crime, technology and pop culture.
ROGER L. MCCARTHY WAS STUCK IN FRIDAY-afternoon traffic on July 17, 1981, when he heard the first radio reports of a catastrophe at the Kansas City Hyatt Regency. In the middle of the hotel's afternoon tea dance, about 1,500 people had been listening to an orchestra play "Satin Doll" when a walkway that spanned the lobby collapsed onto another walkway; both crashed to the lobby floor.
As his Buick Centurion inched along U.S. 101 through Silicon Valley, bulletins updated the story. Hundreds of people seemed to be dead or hurt. McCarthy had no friends or relatives in Kansas City, but news of the disaster held his attention. It was more than morbid fascination -- this was business. He knew his workday would stretch into the night as he responded to the messages for him and his company, Failure Analysis Associates, that were certain to be waiting for him at home.
Indeed, several people had left word for McCarthy, most important among them the law firm representing the engineers and architects who had designed the Hyatt's walkways. "They asked for us," McCarthy says -- not surprising, since Failure, a small, Menlo Park-based firm, employs the world's leading investigative engineers, the intellectual guns-for-hire of the product-safety and product-failure businesses. Called in by the companies and individuals who can afford their top-of-the-line fees, Failure's experts are widely regarded as the people to call when man-made disaster strikes -- and when you need to know why.
McCarthy immediately assembled a team to accompany him to Kansas City the following morning. He'd take a structural engineer, a materials expert and an architect. Unlike risk analysts, who predict how safe it is to fly in a DC-10 or the likelihood of an O-ring failure on a space shuttle, they were to determine what sequence of events touched off the disaster. In this often grim business, fate is an unacceptable answer.
When the Failure team arrived, a rescue crew was scouring the wreckage for bodies. The rescuers had moved quickly through the rubble of concrete, twisted metal and glass, even punching a crane through the front of the hotel. But in doing their job, emergency workers had immensely complicated the investigators' task. Now they had to separate the damage that the rescue operation caused to the materials from that of the "primary" wreckage -- a difficult job, McCarthy says, because "there had been substantial disturbance of evidence."
To understand how Failure Analysis approaches a disaster scene, McCarthy says, it helps to think of the wreckage itself as pieces of photographic film. "In other words, the steel, the concrete, unfortunately the people sometimes, all represent image-absorbing media that form the physical imprints and characterizations on the surface" of the film, he says. "If you know how to read the negatives, you can determine a lot about the picture."
Metallurgists, for instance, can tell by the way a piece of metal is bent the force of energy needed to bend it. "We understand what kind of energies are involved to make the impression. . . . In essence, that's what we do in a failure analysis. We understand the sensitivity of the media." The analysts also reconstruct accidents by designing in reverse -- sometimes putting the found pieces back together again. And they create computer models of the structures to calculate how the materials handle stress.
"Within a day (of the collapse), we knew where the failure initiated and the order of collapse," McCarthy, Failure president and chief executive officer, remembers. But the team wasn't ready to pinpoint a cause.
Meanwhile, though, the newspapers were full of speculations that the accident was the result of "harmonic vibrations" caused by dancers on the walkway. "We were able to discount that theory on the plane going out," McCarthy says. "The natural resonant frequency of the walkway was much higher (than could have been reached that night) -- people can't move fast enough just doing the two-step."
In the hotel that night, the engineers huddled around the blueprints. They realized that the walkway was built differently than it had been designed. "We were surprised and intrigued," McCarthy says. "The contractor had changed the design while building it. I predict he did no calculations. Several engineers lost their professional licenses over it." Failure's team determined that the walkway collapsed because a support rod -- which had attached it to the main building -- pulled through a beam in the walkway. It wasn't the two-stepping that caused the disaster. It was the weight of the people. "That's what initiated the failure sequence," McCarthy says. The walkways were "overloaded, the design was poor and (the construction) was worse." The tragedy occurred in seconds.
WITH MAN-MADE DISASTERS EXPECTED TO STRIKE ABOUT 2,000 times this decade, killing hundreds and causing billions of dollars' worth of damages, the failure business is booming. Failure analysts will go anywhere on the globe to figure out why disasters happened and who's at fault. They're paid exceptionally well -- hourly rates run from $60 to $550 an hour -- to investigate and testify as expert witnesses, often with millions of dollars hanging in the balance. Nine times out of 10, they're hired by the insurance companies or law firms who defend Big Business in lawsuits brought by the Little Guy.
And McCarthy's Failure Analysis teams dominate the field. Failure's experts have been at the scene of most of the headline-grabbing man-made accidents of the past 20 years.
When the supertanker Amoco Cadiz ran aground off the coast of France in 1978, causing one of the largest oil spills in history, Failure Analysis determined why the ship's steering system had failed. When a 1988 explosion on the $1-billion Piper Alpha oil platform killed 167 workers in a fire, Failure Analysis built a model of the platform for a public inquiry by the British government to reconstruct how it happened. That same year, when Consumers Union petitioned the U.S. government to recall the Suzuki Samurai because it tipped over while negotiating the curves of a test course, Failure Analysis rebutted Consumers Union's claims with tests showing that the Samurai driver's action influenced the topsy-turvy performance. The government rejected the recall petition.
And, in 1989, after the Exxon Valdez crashed into a reef in Alaska's Prince William Sound, spilling 11 million gallons of crude, Failure prepared 3-D computer simulations of the accident that may someday be shown at civil trials evaluating Exxon's culpability.
Recently, Failure Analysis dispatched a team of experts to Kuwait to conduct tests on some of its burning oil wells and propose ways to soften the catastrophic effect of the air pollutants. Failure's principal engineer and senior vice president, Subbaiah V. Malladi, who earned a Ph.D. in mechanical engineering from Caltech, has a plan to clean and filter the smoke at the source. He's not sure, though, that Kuwait shares his priority. Kuwait is more concerned with the slower process of extinguishing the flames -- more than 300 fires remain.
Failure analyzes some disasters, such as airplane crashes, with computer simulations; its colorful 3-D animated films might include actual cockpit recordings in the kind of glitzy presentation that could swing the outcome of a trial. Failure says it has also amassed the world's largest private data bases of accident and incident records (240 million), and all U.S. death statistics (every death -- about 30 million -- since 1974). It's gone into the business of selling computer services and software such as NASCRAC -- which uses the principles of fracture mechanics to analyze the behavior of cracks in metals and other materials. A nuclear power plant might use it to help decide whether to replace a pipe that might have a hidden defect.
Failure's computing prowess is so renowned that GM prefers its accident and legal-case data bases over those of the automotive giant's own Electronic Data Service. GM is the largest customer at Failure's 160-acre automobile-safety and test site in the desert near Phoenix. (Work for the automobile industry totaled about 40% of Failure's $69-million revenues in 1991, and GM-related products and services accounted for about $9 million. This past fiscal year, Failure earned $6.4 million with fewer than 525 employees worldwide; this year's financial picture looks even brighter, company officials say. Since June, 1990, Failure has completed more than 200 individual investigations for different clients in which the billings were greater than $100,000.) Failure, which is publicly owned, has experts in at least 25 fields, with nine U.S. offices, plus three in Europe and one in Canada. By contrast, Failure's nearest competitors are privately held firms working primarily in the field of structural collapse.
The company employs more Ph.D.-level engineers (80) than most university faculties, and, McCarthy says, their mission is fairly simple. "We spend a lot of time," he says, "studying how things break and what shape they are in after you munch, crunch, fry, explode and otherwise mangle and totally abuse some piece of material."
BERNARD ROSS, FAILURE'S CHAIRMAN EMERITUS, GAVE A CLUE TO Failure's priorities -- and its almost cocky confidence -- recently on his daughter's 16th birthday. He gave the girl what some would view as a rather unusual gift: his prized convertible Chevrolet Corvair. Corvairs were a major subject of Ralph Nader's landmark 1965 work, "Unsafe at Any Speed," enough to scare off many a parent or potential buyer with its warnings that the car's suspension was defective and the car was dangerous. But Ross says that although he admires Nader's commitment to consumer activism, he believes that Nader's involvement with the Corvair was "totally misleading, uninformed and had no veracity as a matter of technical understanding or engineering. His campaign against the Corvair was entirely emotional -- typical of the way attorneys launch their attacks and crusades."
Nader obviously disagrees, but he and fellow consumer advocate Byron Bloch admit that groups such as Consumers Union can't begin to match Failure Analysis' depth and breadth -- especially on a test track. "Obviously they have a lot of firepower there," says Nader, who calls Failure Analysis a "conglomerate with a huge institutional memory and data bases."
Bloch concurs. "It's somewhat of an unequal contest," says the former KABC-TV consumer reporter who moved to Potomac, Md., to work as a free-lance automobile-safety consultant. "It's an unholy trinity of the auto companies, the insurances companies, plus Failure Analysis. Their resources can be almost overwhelming."
Oddly, Ross agrees with Bloch and Nader that there is an imbalance of power in the automobile-safety-testing business. "It's unfair in a sense to plaintiffs. . . . An avaricious person might say you want to work for big companies because that's where the money is. As an engineer, I've never let anyone hire me to express opinions, which is what happens a lot in this litigation business."
But Failure rarely works for the little guy. Most people simply can't afford to hire them or to sway them from their bread-and-butter loyalties. "It wouldn't be morally right to take a case against General Motors," says John Shyne, 65, a Failure co-founder and principal metallurgist who used to work at Ford before joining the faculty at Stanford. "Failure has many active cases for GM -- we couldn't hardly take a case against them."
"It would be nice, as a citizen of the world, if there was a Failure on the side of Joe Lunch-bucket, but I don't see how you create that," adds a securities analyst and trained civil engineer who helped take the company public. "I don't think Ralph Nader can put together the staff."
Whose side is Failure on, anyway? If the company's engineers were to discover information indicating that a product was dangerous, would they report it to the public, tell the media? The answer, Ross admits, is only if all else failed. "You can't just run off like a banshee and cavalierly broadcast to the world that you've found something," Ross says.
But if the client didn't listen, he cautions, "we wouldn't fall dead. We'd express our convictions and would follow through with our concerns to the highest authorities. It wouldn't be impulsive or emotional. We'd try to work up the ranks of the system. The bottom line is, I've been here since Day One, 24 years, and we've looked at 30,000 or 40,000 accident cases, and we've never had to go to that extreme. There is no product we've looked at that has continued to cause accidents where there's some defect that was ignored."
The system of checks and balances is working just fine, McCarthy says. "Whistle-blowing has become such an accepted activity now that I think every manufacturer operates with an explicit understanding that you've got to operate above-board, particularly if you're publicly held. Even if they didn't, and were the most maniacally motivated people, it's a dollars-and-cents issue -- you can't do it."
SINCE THE DAY IN 1967 WHEN ITS FIVE CO-FOUNDERS -- ALL Ph.Ds -- decided over lunch at a grill near Stanford to chip in $100 apiece to cover future business expenses, Failure Analysis has quietly dominated its field. The company never advertises, instead relying on word of mouth and repeat customers. But its lack of public exposure probably has more to do with the technical nature of the business than the personalities of the engineers who lead the firm.
Failure executives contend that Americans should spend their "safety dollars" more wisely. For instance, McCarthy says it's a mistake to put air bags in cars before current laws regarding seat belts are enforced. Wearing seat belts, he says, increases a person's chances of surviving an auto accident by 40%. Rather than "tax" Americans $500 to $1,000 per car for air bags, he'd first enforce the seat-belt laws, spending the money first to hire extra police to ticket scofflaws, then to encourage other safety measures such as wide-angle rearview mirrors. Air bags, when used with seat belts, increase the chances of survival by only 5%, he says.
Not surprisingly, McCarthy, 42, strongly believes that American industry would be better able to compete with Japan if it were run by engineers rather than by professional managers and MBAs. "There's a feeling that you can hire a guy who is making Pepsi one day to make high-tech computers the next in American industry." His own management philosophy is simple: "We try to make it the rule here that nobody has to work for anybody dumber than they are. That's what we call the legitimate management structure. I think there is a very simple litmus test that you can always apply: What does seniority count for? Here it counts for zero. We believe that seniority is its own reward."
FAILURE ANALYSIS IS EXPANDing its reach beyond the 90,000 accidental deaths that occur each year in America (for reasons other than murder or suicide) to delve into the great unknown -- death by natural causes. The firm has begun to gather data to generate "maps" of environmental toxins and pollutants that contribute to unusually high cancer death rates in various parts of the world. Of the 1.9 million deaths each year in America "that we used to call from natural causes, how many have occurred prematurely because of some sort of toxic, chemical or environmental aggravation?" wonders McCarthy, a Ph.D in mechanical engineering from MIT.
Understanding accidental deaths, McCarthy says, is like mapping the city of Rio de Janeiro. By comparison, understanding death due to natural causes is like mapping the Amazon rain forest. "The Amazon will be mapped," McCarthy says, "just like the genetic structure of humans will be mapped. The only question is when and by whom."
His proposed "map of deaths by natural causes" is more concept than reality at this point, but, hypothetically, it would give lawmakers the facts necessary to regulate the environment. McCarthy says the recent controversy about the pesticide Alar on apples is a prime reason to start mapping. "Look what the accusations did to the apple industry. To say anybody's proven that there are any health risks associated with Alar is ridiculous. But that's part of the Amazon out there. People are raising the question -- is there a health risk? -- but there are no maps."
Americans' declining tolerance for injury, catastrophic illness and death is "a measure of civilization," says McCarthy, whose undergraduate degree was in philosophy. "This society has invested huge amounts of money to make things safer, we've paid a lot of prices that other parts of the world haven't begun to pay. But, as a result, people are less accepting of a catastrophic injury or death. People have demanded more and more answers. Why did this happen, and how are we going to prevent it?"
Failure's own tragedies have reinforced the founders' desire to keep looking for answers. In 1978, Failure's first president, Alan Tetelman, once chairman of the materials science department at UCLA, was killed in a midair collision over San Diego while on his way to investigate a U.S. Navy plane crash. "Alan's death was a real blow to the company," says McCarthy, who had joined the company just three weeks before. "He was the only president the company had known."
In the aftermath, it would have been human nature for Failure to turn
reflective, even to scale back operations -- half of the original 35 employees,
Ross notes, had done their Ph.D work with Tetelman. But the tragedy never
deterred McCarthy, Malladi or Ross from pursuing disasters. The attraction
of puzzling out the mechanics of why things fail is just too strong. "My
real interest is engineering problems -- that's what gets me excited and
turns me on," Ross says. "I'm in it for the engineering thrills."