Comments about technological history, system fractures, and human resilience from James R. Chiles, the author of Inviting Disaster: Lessons from the Edge of Technology (HarperBusiness 2001; paperback 2002) and The God Machine: From Boomerangs to Black Hawks, the Story of the Helicopter (Random House, 2007, paperback 2008)

Thursday, January 14, 2016

Institutional Memory and the Silver Tsunami, Part 2

How to forge a long-lasting, collective memory that leads to safer operations? That's the subject of this followup post to Part 1.

Physical markers can be valuable memory aids. Old stone monuments on hillsides in Japan, erected following long-ago tsunami, warn those who look for them (photo, CBS News): 



Even better are functional monuments, like this building in Banda Aceh that held up against the 2004 event (photo, Daily Telegraph):


Even temporary markers like lockout tags save lives if in conjunction with physical barriers like locks that prevent a valve wheel from being turned, or a blind being unbolted.

The New London explosion – the worst school catastrophe in US history – illustrates the most costly method to build a memory: high-profile, landmark cases that resonates strongly with the public and lawmakers. Soon after, laws were passed requiring odorants in natural gas for sale, and the registration of professional engineers. Also influential were the gas leak at Bhopal, the collapse of the Quebec Bridge in 1907, the Chernobyl reactor explosion, and the Northeast Blackout of 1965.

But even the most vivid memories fade, and the ranks turn over. How to keep them fresh? On July 6, 1988, Steve Rae was an electrical technician aboard the Piper Alpha rig at the time a chain of mistakes led to a natural-gas leak from a high-pressure pipe. 



The chain of events promptly killed 167 men. Twenty years later he took the podium in front of 130 students at a petroleum technician's school to relive the day, its aftermath, and its costly lessons like the importance of a safety-case approach to prevention. "I attended three funerals on the same day,” he told the newly minted graduates, “and that will never leave me.”

Assuming that institutional memory is important, we have to consider this tough question: Will it always make the critical difference? Not alone, it won't. The loss of Challenger seared across NASA and its contractors and made another solid-rocket booster failure very unlikely, it didn't prevent the loss of Columbia seventeen years later.

A common objection to proposals that would fire up a major effort to gather and preserve an institutional memory is that the effort will drain thousands of hours of otherwise productive time, in additional to consultant costs. And once it's done, who'll have the time to go through a mass of recollections that seems less relevant by the year? Won't the competition take advantage of our hard-won knowledge? That's short-sighted, according to Trevor Kletz: “If we tell other people about our accidents, then in return they may tell us about theirs, and we shall be able to prevent them from happening to us.”

I think two broad types of collective memory are achievable and worthwhile in high-risk industries, each in its way. And they don't have to be a time-burner.

The two types are motivational memory and working memory.

A motivational memory is less about technical details and more about remembering the need to work cooperatively and safely. 

Why do newly graduating structural engineers in Canada join in the ritual of the Iron Ring? It's not a refresher on statics and dynamics, it's a reminder that people die in collapses if experts don't sweat the details. Jack Gillum has given speeches about the catastrophic collapse of walkways at Kansas City's Hyatt Regency Hotel in 1981. Gillum, as the engineer of record, was found negligent in not catching a fatal flaw in revised shop drawings. He lost his Missouri license over it and 114 people lost their lives that night. Many more were injured in the collapse. A firefighter had to perform an amputation with a chain saw. I heard Gillum speak at an engineers' forensic convention fourteen years ago, and what he said that day remains with me. Further, I believe that when employees are injured on the job, managers who controlled the job site are obligated to visit them in the hospital, and attend funerals too.

Working memory: rather than taking aside all employees for long recorded interviews as they approach retirement, consider strengthening the day to day, functional memory as held in the minds of high-performance teams. Confronted with the need to design a new line of cars from scratch, Chrysler split the job among one hundred “tech-clubs,” each responsible for a key component or assembly. By forcing early companionship between design engineers, marketers and suppliers, Chrysler found it could speed development and cut costs. One advantage of a team approach is that expertise is broadly distributed, lowering the risk that a single employee's departure could cripple a critical operation. At its best, that's how the American military works, putting hugely consequential decisions in young hands, mentored by old hands.

Another argument for taking a team approach is that a team is, or can be, much more than the sum of its parts. According to psychologists who study memory formation both individual and collective, people remember an incident most vividly if they've participated in a group that discussed it afterward. Safety-oriented tailgate talks at jobsites are a good time to bring up lessons learned, fresh off the docket.

Group discussions about accidents and close calls also build up the motivational memory. Through such discussions, even people who weren't at the scene of an explosion feel the emotional impact, and it inspires them to go the extra kilometer. As Yogi Berra might have said, no one wants to experience disaster déjà vu all over again.


Saturday, January 9, 2016

Institutional Memory and the "Silver Tsunami," Part 1

If I had to think of one topic that comes up during every public appearance I make, across all industries and specialties, it's the urgency of replacing the baby boomers now heading for retirement: preserving not just the technical skills, but the memory of what works and what fails. 

This story comes to mind. Two months after a wave from the Great Tohoku Earthquake demolished hundreds of towns in northeast Japan, the Washington Post described one that survived: Fudai, a community of 3,000 nestled in a narrow valley that was wide open to the sea. In 1972 its mayor called on the town to build a 51-foot-high floodgate. The project attracted much opposition over the cost ($30 million) and the land required to hold off a big wave -- the next big wave, in the view of then-mayor Kotoku Wamura. 

As a young man he'd seen the aftermath of a 1933 tsunami that killed over 400 in Fudai alone. As mayor, he led the project and faced a lot of hostile questions: Why did the town need it? Why so high? Other Japanese cities had put up gates and seawalls, but none were so high. How could Fudai pay for it?

Wamura was undaunted. A good thing, too: When that next wave arrived on March 11, 2011, water lapped over the top but the damage was inconsequential; the only death was one man who had climbed over to check on his fishing boat. Without Wamura's big wall, Fudai would have been reduced to bodies, trash, and rubble. Again. 

Memory – vivid and awful – carried Fudai's floodgate project forward against all opposition. It needed more than the mayor's individual memory: it was a collective memory of everybody old enough to have seen the effects of the 1933 wave.

The subject of memory and how to hold onto it is a hot topic because the baby boomers aren't babies anymore. Experts warn that the looming retirements, across all sectors of the economy, is a “silver-haired tsunami.”

However much fifty- and sixty-somethings look forward to retirement, they're equally eager for anti-Alzheimer nostrums, whether vitamin packets, red wine, Soduku puzzles, or online memory tests. Worries over memories that slip-side away extends to the largest scale. Consultants are wagging their fingers at companies and agencies like NASA, warning them to capture their “institutional memory” with extended videotape interviews and copious databases. 

They're referring to the unwritten knowledge held by skilled workers, seen-it-all foremen, and hands-on managers. It's trouble-shooting. It's the agility that strikes a balance between handling existing projects and taking on new challenges as conditions change. In short, it's the know-how that gets things done and heads off the ICE, the Imminent Catastrophic Event.

Before looking into what collective memory is, let's think about individual memory. While our brains are sometimes compared to a computer's storage banks, people are radically different from computers in how they collect and store information. In 1861 Abe Lincoln referred to the mystic chords of memory, and he wasn't far off the mark. Memory is not a predictable set of nerve connections. We know more about how it goes away than why it stays.

Experts in mnemonic techniques assure us that with training and jaw-aching concentration just about anybody can erect a memory palace in their minds and then wow their friends by quickly memorizing the order of an entire, shuffled deck of cards. Meanwhile, most of us still have not a memory palace but something more like a drafty house. Even without the affliction of Alzheimer's, facts blow out the back door when we're not looking, and other facts get mixed up like old keys tossed into junk drawers. Check out this “Jaywalking” episode from Leno, for a wacky stroll through history as feebly recalled by the man on the street. 


The good news is that humans are, or can be, quite good at building and holding a body of knowledge. Knowledge is what drives our decisions. It's a combination of skills, recalled facts, and insights, and is unique to each person. 

Recall my drafty-memory-house analogy? Now imagine a snug, warm greenhouse in the back yard, a place for plants to grow and thrive. For an amazing example of how people can amass huge bodies of knowledge when they must, check out Mark Twain's Life on the Mississippi


He describes how each licensed pilot of the 1850s had to know the channels suitable for big steamboats along more than a thousand miles of unmarked river, storing the images for use by day, night, and in the fog … and then absorb new information as channels, snags, and sandbars changed.

Such vast collections fit into a few pounds of brain tissue because they're braced and motivated by personal experiences, vivid stories from trusted sources, reading, and certification courses. 

That's memory and knowledge at the micro level. What about macro: Can an entire company, or even the workers across a single plant, share a “collective memory”? Safety expert Trevor Kletz, author of What Went Wrong? and Still Going Wrong, believed so. 

The tendency of refinery and chemical plants to lose their institutional memory of past disasters, about every ten to fifteen years, has been a concern in the chemical-processing safety literature for years. Writing in Modern Railways, Roger Ford said that accidents happen “when the last man who remembers the previous disaster retires.”

On the other side of the memory-is-good question are advocates of extreme makeover, corporate style. If what Robert McMath calls Corporate Alzheimer's is the collateral damage, so what? To these skeptics, it doesn't matter whether anybody in the organization recalls past problems and how to avoid them, because the key is going forward. Here are their arguments:

Didn't the fabulously successful Henry Ford say in 1916, “History is more or less bunk.... We don't want tradition. We want to live in the present, and the only history that is worth a tinker's damn is the history we make today.” 

Here's how Ford's thinking lives on:

  • “All damage and injuries are due to either (1) unpredictable flukes of fate, never to be repeated and therefore needing no attention, or (2) errors by low-ranking workers who recklessly flaunted their training and operating manuals. So there's nothing to learn.”
  • “Internal histories that capture damage incidents, close calls, and lessons learned would be expensive to assemble, and then plaintfiffs' lawyers might get hold of it, so why go to the trouble? It's better to plead ignorance after the next bad headline, and do it convincingly.”

Now for the other side of the coin. There's a museum called The Collection at New Product Works of Ann Arbor, Michigan, for which people pay a lot of money to tour. The shelves hold more than a hundred thousand products, most of which you can't find anywhere else because they flopped so quickly, like Look of Buttermilk shampoo, Male Chauvinist Aftershave, and a urine-colored bottled tea called “Tea Whiz.” 

Even the museum at Ann Arbor, big as it is, can't display all the ways that firms and governments forget at least as much stuff as we mere humans do. And about as fast. 

The problems of rapid employee loss and turnover are magnified by the loss of supervisors with long and plant-specific experience.It's been said that foremen and supervisors act like synapses of our brains. On the job, they link individuals into functional units that span the organizational charts; along with motivated higher-ups, they can press for prompt action to head off a disaster. Critics like Kevin Foster call the discharge of such experts not downsizing but dumbsizing. But the nation would still have a memory drain problem even if companies reversed direction, because there's a graying workforce that is sure to move on sooner rather than later. Mack Truck built an assembly plant for the Soviet Union, but when the opportunity came to win a contract to refurbish the facility in Russia, Mack lost the bid to another company because the company experts on the original plant had moved on and no working memory remained of how, or why, the truck plant was laid out. 

A plant doesn't have to be halfway around the world to turn into something dangerously unfamiliar, as employees change jobs and memories fade. Disaster annals are full of spectacular events triggered after an incoming worker looks at some pre-existing gizmo, decides it's getting in his way or slowing him down, and changes it without asking anybody. This can be a enormous hazard at an oil refinery, where an peculiar-looking vent stack might be essential to avoiding a vacuum that would cause two chemicals to react and mix at the wrong time. In a perfect world, a complete set of plans would not only show the machine in its actual, “as built, as modified” status, it would also have little tags explaining what the tubes and safety valves in a boiler room or refinery are there for, in case someone has the hankering to tinker.

==

How to forge collective memory that leads to safer operations? That's the subject of Part 2.