Seven Practical Steps to Build a Culture of Safety and Human Reliability

There isn’t one magic bullet, but rather a ‘consolidation of subtleties’

Published: Monday, November 21, 2022 - 13:03

A technician spills a toxic chemical. She isn’t injured but easily could have been. The hazmat cleanup costs more than $10,000 and shuts down a critical building for a week.

An electrical engineer flips the wrong switch in a substation control room. He isn’t injured. But within seconds, a $50,000 transformer is destroyed.

Three financial clerks in two different countries are processing payments for a large bank. They intend to schedule a routine $8 million payment. Antiquated software makes errors hard to catch. The clerks accidentally wind up sending $893 million instead.

From talk to action

Talking about building a culture of safety and human reliability is easy. But how many great ideas are talked about and never actually get put into practice?

The real skill is to be able to transform good ideas into practical steps that you and your people can apply immediately. In this article, that’s what you’ll get.

There’s no one secret or solution. Instead, many successful companies around the world have built a culture of safety and human reliability using a “consolidation of subtleties”—a combination of seven practical steps such as these.

1. Take a learning-based approach to errors

Many leaders see errors as failures that can be eliminated with more rules, thicker procedures, and stricter punishments for people who don’t comply. If you’re in a work culture that’s stuck in this old-school, control-based approach to errors, then consider labeling it that way: “Hey, are we stuck in a control-based approach as we’re discussing Tuesday’s incident?” The more you label it, the more you’ll be aware of it, and the less you’ll be stuck in it.

When ready, propose the alternative—the learning-based approach. If an assembly line breaks, it might be easy to find a single root cause, such as a faulty bearing. But with complex jobs, when a team of experts misunderstands each other, the search for what went “wrong” often turns into a game of assigning blame. The more you learn about how those experts actually do that job, the better you can identify real-world improvements that fix the problem without fixing the blame.

In their paper, “From Safety-I to Safety-II,” health research experts Erik Hollnagel, Robert Wears, and Jeffrey Braithwaite write that “It is necessary to understand how such everyday activities go well—how they succeed—in order to understand how they might fail. From a [safety] view, they do not fail because of some kind of error or malfunction but because of unexpected combinations of everyday performance variability.” So, after the next incident or unwanted error, don’t start by asking, “What went wrong this time?” Instead, start by asking, “How is it this job got done right 99 percent of the time?”

2. Create psychological safety

It’s easy to destroy and challenging to create—yet research from Amy Edmonson at Harvard and Google’s Project Aristotle reveal that psychological safety is key to successful, safe, engaged, and reliable teams. When a serious incident occurs, the team involved often fears the judgment and blame that traditional investigations can create. So they say as little as possible. Some investigators respond with threats of sanctions unless they get “the truth.” In standoffs like these, psychological safety evaporates, leaving only fear and distrust. To build psychological safety after an incident, don’t start with a judgment like, “Joe failed to do [X].” Instead, ask in good faith, “What did Joe do, and why did it make sense for him (at the time) to do that?”

3. Lead after-action reviews (AARs)

For more than 30 years, these psychologically safe, semistructured, post-job team debriefs have been used by an increasing number of high-hazard industries worldwide. After your next successful, complex project, instead of asking, “What could we have done better?” ask these four questions that were initially developed to accelerate learning in the U.S. Army:
1. What did we set out to do?
2. What did we actually do?
3. How did it turn out that way?
4. What will we do differently next time?

4. Transform investigations

Traditional investigations often “name, shame, blame, and retrain.” The result? Fear, silence, and box-checking on corrective actions. The alternative? Instead of asking, “What was the error, and who made it?” ask, “How did our processes set that person up to make that error? And how can we improve our processes to set our people up for reliability and success instead?”

5. Apply defenses

Peer checks. Three-step communication. Checklists. These and other simple yet powerful defenses have proven successful for decades. You can learn them in a few hours and get real-world results immediately. Pick a job that you and your team regularly do. Consider writing or updating the checklist for that job to include only the three to seven items most often missed. One physician from Baltimore helped save 1,500 lives in 18 months with this classic defense.

6. Improve systems

Instead of trying to “fix” your workers, improve your work processes and systems. How? Pick a process that you regularly do. With trusted frontline experts, brainstorm one low-cost, low-risk, low-fear, low-maintenance process improvement that would make it easier for the experts to do the right thing in that process. For example: Companies with fleets of trucks have dramatically reduced serious injuries and saved millions of dollars each year by simply avoiding left-hand turns.

7. Build resilience

The world’s most high-reliability organizations (HROs) don’t try to eliminate all errors. They don’t proceduralize everything, either. Instead, they build resilience so that most errors become easier to detect, recover from, and learn from. How do HROs do this? One method they use is to look for “weak signals,” like the sound an engine makes when it’s just starting to develop a problem. Novices miss weak signals in processes. But experts sense them and act quickly to manage errors before they cascade into catastrophes. So talk with one or two trusted peers and identify one weak signal for a complex job that you all do. Name that weak signal, and what to do about it. Then teach that to your apprentices instead of hoping that they’ll discover it on their own.

If these seven steps seem like a lot, don’t worry. Just pick the one that resonates with you the most right now, and discuss it with a few trusted colleagues. When ready, try it out in a low-cost, low-risk micro-experiment. A few small, quick wins will help you build momentum fast.

About The Author

Jake Mazulewicz

Jake Mazulewicz shows leaders in high-hazard industries why errors are signals, not failures, and how to address the deeper problem so everyone can work more reliably and safely. A keynote speaker and advisor all across North America, he has a decade of experience in safety for electric utilities and has served as a firefighter, an EMT, and a paratrooper. To learn more, visit www.reliableorg.com.