PAPod 619 - Normalization of Excellence: Turning Safety into Strength

0:13 Hey, everybody. Welcome to the Pre-Accident Investigation Podcast. I'm Todd, your host. Howdy, pumpkins. I hope you're doing well. So in North America, we're turning into fall. And in New Mexico, the leaves are changing. There's color in the mountains. Gold, red, green, brown, always brown. We got lots of brown. We're good with the brown.

0:34 But yeah, it's a seasonal change, which is a part of it. It's the circle. You know, we just keep spinning around the sun and life just keeps happening and we get to spend a little bit of time together each and every stinking day. So I have so much to talk about this week. It's crazy about this week. I want to start with the gang at Southwest Airlines.

1:00 And so I should tell you that Chris, who is a remarkable safety person and super interesting in really operational reliability. I would say those guys pretty much have that going on. They opened up a day where I could go down and watch a class that I'll just be really honest, I kind of begged to watch.

1:23 So do you guys remember, and even if you don't remember, and I think it's 96, I'm going on memory, a better person would to look this up. Diane Vaughn, who's a sociologist, we're going to talk a little bit about Diane Vaughn, wrote a book about the challenger accident. And she coined a phrase, the normalization of deviance.

1:45 And what she talked about was that slow drift into failure. That's a term we've heard. Decker's got a book entitled that. And we all know if you look at black line, blue line, the blue line corner, all this sort of represents this drift into failure. And I actually think this notion of the normalization of deviance is pretty important.

2:07 It's interesting because deviation itself is quite normal. It happens all the time. When we normalize the deviance, which is why she uses that phrase, what's happening is we take slight deviations, you know, minuscule little drifts, little paso-y-paso, step-by-step drifts, and they become the new normal.

2:30 Now, you could ask me, and I got asked in a workshop just this week, how many times do we have to react to something abnormal before it becomes normal? And I actually think that's a really good question. And I'm afraid the answer is going to be not very many times. But if you're out there doing research, Matthew Hallowell, I'm talking to you. This could go well.

2:52 That would be something really interesting to look at. Because we do have this ability to drift into failure. And Diane Vaughn writes about it quite elegantly. And we've picked up on that over the last 30 years or so. And the Challenger event, because there was much to teach us in the Challenger event.

3:09 And we've talked about it. And we've used it. And I've actually think the idea has made us better, but that's not what I went to see in Dallas at the Southwest guys, cool headquarters that looks like a spaceship or a set for, I don't know, Battlestar Galactica. It's pretty cool building, right? They have a class called the normalization of excellence.

3:32 Oh, I got your attention right then, didn't I? Yeah, you just perked up. Yeah, I saw it happen. Don't act like it didn't happen. And I was really curious about that. And one of the things I think I'm going to do, and if you think this is a good idea, just tell me, is we can bring some of those guys on and talk to them.

3:51 Because what they did was basically take the four components that Diane Vaughn introduced in her work, that leads to this normalization of deviance and just reversed them and said, well, what if we looked at the negative as a positive? What if we phrased it as a positive? Wouldn't that also then allow us to normalize towards excellence?

4:16 And that's what we went to do. And it was great. I met so many great people. There was a gang of folks there. We just had an amazing time and I can't thank them enough. I also think it's the beginning of yet another conversation that we're going to have in the safety world. Because I do think this shift towards productive safety is a really important shift that's happening.

4:40 And it's happening all over, no matter where you are, what you do, is this idea that somehow safety's job was to reduce risk is a really important foundation. It's something we stand on. But I actually think safety's real job is to make work better. And more reliable, and more interesting. And so, that's what we did. But I mean, it was crazy. Like, the guy in charge of maintenance spent the day with us.

5:08 I mean, that's saying something, right? I mean, that's saying the importance of this discussion. And it was great. I mean, we just had a lot of discussions. We told stories. We learned from stories. We learned from each other. Mostly, it's just really good to sort of check in on one another, which is kind of what I always think the podcast does anyway.

5:28 I mean, it's a, it's a big part of it. Then I got to spend the day with the plant managers for Vista Energy with, with Charles Major and the gang. And that was just amazing as well. I mean, it was a really good couple of days. Then I got to hang out with the propane dudes because you can't spend enough time in propane and propane accessories. And that was a super fun day.

5:48 So I would say, you know, all in all. I had a pretty productive week and my flights were only like, you know, a couple hours late each time because that's what happens to me because I think I'm being punished. I don't know. This, this idea of the shift in safety that we've been dealing with for the last 20 years or so, it's really starting to make more sense operationally.

6:15 And I like it when stuff moves from the academic side into the applied side. Into the operational side. And that's kind of what we're having the conversation about, is think about the principles of human performance. You know, we talk about those all the time. Those at one time were sort of heady and very philosophical.

6:38 Now they've been kind of operationalized. And you see people genuinely talk about the fact that if we create an environment where we can't talk to each other, we're going to squelch learning. And if we squelch learning, we're going to absolutely squelch improvement. And so that challenge to make safety a productive part of the organization that makes doing its mission more effective.

7:05 That's probably exactly the right conversation that we should be having. And I think that's a very, very interesting thing to talk about, which kind of leads to today's podcast. So today, I don't know what to do with this because, you know, we did that big discussion on AI and it was so good to talk to Brent and he's always interesting. I mean, he's interesting.

7:24 One is he's got a New Zealand accent, so that's just interesting right away. But also the struggle with sort of figuring out where AI fits in the world is a pretty important struggle. And so because we did this pod around this AI thing, something really unusual happened. And that is a person kind of anonymous. I don't really even know who did this, did an AI and created an interview.

7:52 I'm guessing they used Notebook LM, but I don't know. I mean, it sounds like those guys. And basically the question they ask, I think, because I listened to it, you're going to hear it too, is how have we matured on the journey over the last several years in our thinking around human performance and its impact on operations?

8:14 Kind of exactly what I went to Southwest Airlines to observe. A normalization of excellence. Now, I'm not saying I'm excellent or we're excellent. I mean, that would be pretty heady and kind of stupid. But I do think this idea of asking this question was interesting. And somebody thought it was interesting enough to actually produce this little interview.

8:36 So that's what I want to do today is share with you what that interview kind of talks about and see what you think, because I'm curious what you think. I don't really have a lot to do with this. I mean, I'm not even sure how it happened. Well, I am sure I don't know how it happened, but I'm not even sure how you'd make this.

8:57 But I think it's very it's very forward thinking to use AI to address this question. And I'm guessing it's a supportive move towards some uses of AI that don't involve surveillance. At least that's what I sort of drew from the invitation that was attached to this file when I got it. And I'll grant you, I'm super worried about the surveillance side of this, but I'm worried about it for a lot of reasons.

9:24 Safety for sure. Reliability for sure. Justice for sure. But I'm also worried about it in kind of everyday life. It just seems like, I don't know, people are a little too much in my business, if you know what I mean. And I think you do. So let's sit back and listen to this and tell me kind of what you think about this interview.

9:44 I'm curious to see if you find any value in it at all. I'm curious. I don't know. The jury is definitely out from my side. And so I think the very best thing I could do is take it to you and ask you guys what you think. So without any further ado, from an anonymous listener who clearly is motivated and cares, here's an interview that was created on my journey...

10:09 Over the last several years and how my thinking has changed and matured. I'm not sure it has. That's a complete different podcast. But let's listen to it and see what you come up with. Here we go. Welcome to today's Deep Dive, everyone. We are jumping into some incredibly fascinating territory today, really exploring the evolution of workplace safety and human reliability. Yeah, it is a huge shift in how we think about work. It really is.

10:42 And to do this, we are pulling from two major works by Dr. Todd Conklin. So we have excerpts from his Five Principles of Human Performance and also from Better Questions. Right, which are both just, I mean, completely foundational if you are looking at this stuff. Absolutely. And our mission today is really to trace how Conklin's work has matured over the years.

11:01 You know, going from these sort of early foundational philosophical truths about human error all the way to these really mature practical frameworks. Frameworks that, frankly, just completely upend how modern organizations respond when things go wrong. Exactly. And honestly, whether you are managing a huge corporate team or you're designing a new process or, I mean, maybe you're just trying to understand why you keep making the same mistake in your own personal life.

11:27 This is for you. Yeah, it changes everything. It fundamentally changes how you view human error. Okay, let's unpack this. So to really understand how far Conklin's work has come, you kind of have to look at the baseline he was fighting against in the beginning. Right, the old way of doing things. Exactly.

11:42 For decades, the holy grail of reliability in business was this concept called behavior-based safety. Oh, yeah. And the whole underlying assumption there is that workers are basically just like cogs in a perfectly choreographed machine. Right. And if a gear slips, the machine breaks. Precisely. So the logic for management was that accidents are simply the product of bad choices.

12:02 If we just get workers to behave better, you know, reward the good behavior, punish the bad, then we achieve perfect reliability. Which is such a seductive illusion, right? I mean, it takes this incredibly complex thing, human psychology, and just turns it into a simple spreadsheet. It really does. But the sources point out that this is so flawed.

12:21 Because if managing human behavior was really as simple as treating people like simple machines or like vending machines where you put in a reward and get out a safe outcome. We wouldn't have accidents anymore. Exactly. We would have eradicated workplace accidents back in the 90s. Yeah, because people just aren't machines. And that was the early pivot in Conklin's work. He realized that workers, they do not show up to their shifts intending to fail. Right.

12:45 Nobody wants to get hurt. No one. They come to add value. And the most crucial takeaway from this early period is that the exact same behaviors that create success are often the exact same ones that lead to failure. Oh, wow. Wait, explain that a bit more. Well, the only difference is how the system responds when a disturbance happens.

13:04 Workers are always improvising and adapting to keep production moving. When that improvisation works, they get a bonus. When it fails, they get fired. Right. The behavior didn't change, just the outcome. Exactly. So as his thought matured, Conklin really distilled this into his five principles of human performance and the bedrock of that. I mean, principle number one is just simply that error is normal.

13:27 Which sounds so obvious, but it is revolutionary for a corporation to admit. Totally. You can't train or wish or punish error out of the human condition. And then principle two is that blame fixes nothing. And principle four is that context influences behavior. Context is huge. It is. And there's this brilliant analogy in the text, the airport analogy, which I just love. Oh, yeah. That is a great one. Right.

13:51 Because you think about a place like LAX or London Heathrow. Millions of people. And they are stressed. They're exhausted. They are looking at their phones, dragging luggage. It's pure chaos. Total chaos. But they rarely bump into each other. And it's not because everyone makes this deeply conscious choice not to hit anyone.

14:08 It's not a behavioral choice. No, it's because the context, the actual physical environment, is designed to absorb human inattention. They have wide spaces. They don't have, like, random unmarked stairs in the middle of the hallway. What's fascinating here is that it introduces this concept of local rationality.

14:26 Local rationality. Yeah. Okay, break that down. So local rationality basically means that workers do what they do for a reason that makes complete sense to them at the time. Oh, okay. Yeah. Given the specific pressures and context they are in, their actions are rational. So when an accident happens, blaming them retroactively just makes the error look like a choice. Which it wasn't. Right.

14:46 Blame just sucks up all the emotional space in the room without offering any actual operational value. It doesn't fix the hallway. It just punishes the person who tripped. Exactly. But, you know, shifting away from that blame is really hard. And the text talks about this real-world friction Conklin faced as his theories matured.

15:05 Oh, the resistance was massive. Yeah, it challenged deeply held beliefs. There's this story in the sources about a guy known as Captain. Captain, right. Yeah, he was this highly respected safety leader. A guy who, like, literally walked around hugging his co-workers. He cared deeply about his team. A really good guy. Right.

15:23 But when Conklin started saying that we have to stop blaming workers for systemic failures, Captain just got furious. He stood up in this training session and basically shouted that you just have to admit that some people are stupid. Wow. Yeah. He said like 10% of workers are just idiots and that's why they get hurt. And then he walked out of the room.

15:41 Which is such a defensive reaction, you know? It really is. And it ties into this provocative thought experiment from the text about drunk driving. Here's where it gets really interesting. Is drunk driving a choice or a mistake? That's a heavy question. Because if you decide it's a malicious choice, your entire safety strategy relies on punishment.

16:01 You increase fines, build jails, and you anchor everything to the hope that fear will force an impaired person to make a good choice. Which is completely illogical because an impaired brain literally can't make reliable decisions. Exactly. But if you view it as a mistake or a system failure, you design interlocks.

16:22 You install breathalyzers that make the vehicle impossible to start. You design the failure out of the system. And that synthesizes the whole point perfectly. Because the consequence of a failure is completely agnostic to the cause. Meaning what exactly? Meaning a broken leg doesn't care if the worker was smart or stupid.

16:41 A severed finger doesn't care if it was a genuine mistake or an intentional shortcut. The consequence is identical. Oh, I see. So calling a worker stupid, like Captain did, is not just mean, it's actually a lazy investigation tactic. It kills curiosity. Exactly. It stops the organization from learning anything about the system.

17:01 Because the minute you say he was just being stupid, you stop looking at the broken machine or the terrible schedule. You're off the hook. Management doesn't have to change a thing. Which leads us right into how Conklin's work actually transitions from just this philosophical theory into real operational practice.

17:17 Right. Because it's no longer just about saying, hey, error is normal. It's about how you actually access the knowledge from the floor. And this is where we see the shift from traditional investigations to what he calls learning teams. Yes, learning teams are crucial. Traditional investigations just look backward to assign blame.

17:36 Learning teams look at how work actually happens on a normal day. Right. And there's this great analogy in the source about burnt eggs. So imagine someone is cooking breakfast and suddenly smoke is everywhere. The eggs are completely ruined. A disaster. The breakfast is ruined. And you have two choices.

17:52 You can either just throw the pan in the sink and go out to eat at a diner, which gives you immediate satisfaction. You get fed right away. Right. Or you can take the time to figure out exactly why the eggs burned. Was the heat too high? Was the pan bad? Which takes time and effort. Exactly. The pressure to just fix the problem by going to the diner is always stronger than the pressure to actually learn why it happened.

18:14 But fixing without learning is totally useless. Because if you just go to the diner, you're going to burn the eggs again next week. Because you never fix the stuff. Exactly. And if we connect this to the bigger picture, the text uses this historical medical analogy that is just spot on. Oh, the bloodletting one. Yes.

18:33 Hundreds of years ago, doctors used bloodletting to treat everything, right? And they did it because they were completely convinced they already knew the answer. They thought they had it all figured out. Right. And now we look back and realize that their certainty was the problem. The mature approach is realizing that not knowing is actually really powerful.

18:53 Conklin says, the answer is the enemy of the question. The answer is the enemy of the question. That is so good. It means that by admitting we don't actually know how the work is being done on the floor, leaders open the door to true organizational wisdom. You have to ask the workers. And to really see what that looks like in practice, the text gives us this incredible real-world case study from Bob Edwards.

19:16 Oh, the receiving doc story. That is a powerful one. It really is. So let's walk through this. There's a near-fatal incident on a busy receiving dock at a plant. A worker driving a fork truck misses hitting a pedestrian co-worker by literally just inches. Terrifying. An absolute near miss. And in the old behavior-based model, management would just fire the driver, right? Or punish him, tell him to be more careful and move on.

19:40 Yeah, put up a sign that says, watch out for pedestrians. Exactly. But instead, this management team used a learning team. They brought five workers together around a flip chart, and they just asked them to map out their morning routine. No blame. Just tell us how the work actually happens. Right. And that was the aha moment. because they discovered that at exactly 9.30 a.m.

20:01 Every day, there were three completely conflicting activities happening at the exact same time. A total systemic failure. Right. You had plant fork trucks arriving to get supplies. You had external delivery trucks trying to unload. And this is the craziest part. 9.30 was the floor workers' break time. And they had to walk right through the dock to get to the break room. Yes.

20:22 To go to the bathroom and the break room, they had to walk right through this chaotic zone of heavy machinery. So what's amazing there is how this mature approach yielded system-level solutions, right? They didn't just give the driver behavioral scolding. No, they actually changed the environment. Exactly. They moved the parking lot for the trucks, they painted a protected green walkway for the pedestrians, and they staggered the external deliveries.

20:46 They changed the context. But the most profound detail of that whole story to me is what the plant manager did with the driver. He apologized to him. He apologized. The manager apologized to the driver for the terrible, dangerous environment they had forced him to work in. Wow. Yeah. And because they treated him with that kind of respect, that driver eventually became a safety lead for the whole plant. Because they realized he wasn't the problem.

21:11 He was the person trying to solve the problem in a bad system. Exactly. And that just perfectly summarizes the whole maturation of Conklin's work. Going from that basic understanding that, hey, human error is normal to fighting that toxic, you can't fix stupid culture. Right. And moving all the way to deploying these learning teams that actually co-create safety with the workers. Yeah. The people doing the actual work.

21:33 So what does this all mean for you listening right now? I really want to challenge you to evaluate your own response to failure. It is a hard habit to break. It really is. When a system breaks down around you, do you ask who failed? Or do you pause and ask, how did the system fail? You really have two choices.

21:54 You are either getting better or you are getting even. Getting better or getting even. You can't do both. And I want to leave you with a final thought to ponder today. If hazards are not static, permanent things, but rather these fluid variables that constantly move in and out of our lives based on context.

22:14 Which they absolutely are. Then how much of our daily stress is simply caused by stubbornly trying to manage human behavior instead of just redesigning the actual pathway we walk on? Think about that the next time you find yourself blaming someone for it. Hmm. What do you think? I don't know. I think it's, uh, it's definitely interesting.

22:42 The technology is kind of amazing. It's, they sound like real people. Like you could meet them. Like, I don't know that they might like one of them might even have bad breath. I mean, that's how realistic it sounds. Thank goodness. We don't know that answer, but it's there. And the content seems interesting. I mean, I don't know.

23:00 I don't even know if I agree with it, But that's a great little addition that somebody sent as a response to the conversation we had about AI just a week ago or so. Excuse me. Sorry. I had some incredible enchiladas earlier, and now I'm enjoying them one more time. Kids, growing old is not for the weak. Let's just say that.

23:25 So that's the pod for today. I hope you enjoyed it. Have a great, great, great week. I hope good things happen to you. Look for me. I'll be on the West Coast saying hi. Kissing babies. You know, it's all going on there. Until then, learn something new every single day. Hope you did today. Have as much fun as you possibly can. Be good to each other. Be kind to each other.

23:44 Take care of each other. Check in on each other, even the least among us. I think it matters more now than ever before. Have I got a story for you? I'm trying to still think about how to tell this story. It's that good of a story, but I don't know how to tell it. I'm going to have to figure it out. Until then...