EP 628 - Sean Brady

Safety FM with Jay Allen

Speaker 2

00:08

Okay, we are broadcasting live because that's what we've been doing this week. Of course, having all kinds of fun times as we are hanging out at Energy Safety Canada in Banff. Yes, you know, that lovely location.

S2
Speaker 2
00:21
So let's talk. Sean, directly off the stage, we have you in the chair in front of us, probably 1 of the best presentations I think I've seen in my whole life and I'm not saying that because you're really good.

S1
Speaker 1
00:30
Thank you very much,

S2
Speaker 2
00:31
Jim. So talk to me because, okay, you covered this very intriguing incident with Boeing. Well multiple incident was Boeing How did you come up with the concept and why was it so important to you?

S1
Speaker 1
00:43
I Was fascinated by Boeing because well if I step back a bit I look at a lot of failures, investigate failure, but I look at a lot of famous failures as well. And Boeing is 1 of the best examples of a failure where you have a front line incident, 2 plane crashes, But the seeds of that failure are sown way back both up the corporate ladder inside the organization and back in time. So it's a very time-dependent failure that the company made decisions that locked them on a path that not made failure inevitable but increased the probability of problems.

S1
Speaker 1
01:16
Now we see other organizations do the same and I think for me we spend a lot of time talking about safety, the frontline, empowering them, that's incredibly important but there's this other side to it which is further back up the line. A lot of the problems the frontline have to deal with are set by people who don't necessarily intend to say you know cause problems but what they end up doing is create an environment where those problems essentially emerge out of it so for me it is that type of fear that very you know clearly shows that path dependence from from management decisions.

S2
Speaker 2
01:51
So so if because there's going to be people that are listening to this that were not here. Could you do a brief overview of what your presentation was about? And then we'll kind of dive into some of the more detailed information if you don't mind.

S1
Speaker 1
02:04
Yes, so Boeing comes out with the 737 MAX and within 5 months there's 2 crashes and 346 people are killed. After the first crash it's very much blamed on human error as is done in many of these cases that it was the pilots and the maintenance crew of the airline. What would transpire is that it was actually a problem with the plane and what had happened was the plane had a software system on board that could take control of the plane at certain points in the flight.

S1
Speaker 1
02:35
It relied upon a single sensor, the sensor which is an angle of attack sensor on the front of the plane that measures the angle of the plane essentially as it's flying. If anything goes wrong with that sensor, then it feeds erroneous data to this MCAS software system. The MCAS software system can take over the plane and cause serious problems, as it did in these 2 situations with the 2 faulty sensors. A lot of the presentation is really talking about the history of what happened in Boeing to produce a situation where all it took was a single sensor failure to crash a plane.

S1
Speaker 1
03:06
Because you know, in aviation, we have layers and layers of redundancy. So how did we get to a point without that redundancy? And it really goes back to, you know, Boeing was an engineering first company from 1916, by 1989, 140,000 people working in it. The belief was you build great planes and you'll be financially successful.

S1
Speaker 1
03:26
They had a great line they used to have in Boeing which was, you know, we hire engineers and other people. Engineers were encouraged to demand what they needed on their planes to have them safer. So that was the culture but then in 1997 they merged with McDonnell Douglas, very different type of firm, cost-cutting, return on shareholder investment was everything and over the next 10 years that the whole company morphed into a company away from engineering to a company where the money was was was king essentially and cost cutting was was everything in returning shareholder investment was everything And what you see as part of that is that when Boeing were responding to the challenge of Airbus's new A320neo, they had to work faster. They had to work within this new culture they created.

S1
Speaker 1
04:11
They started to say well Let's not build a brand new plane. Let's take an existing 737. We'll call it the Max we'll put more fuel efficient engines on it, but most importantly is we'll be able to say to the airlines it flies exactly the same as the old 737. Really important, because if you can say that to an airline, you can basically take their existing 737 pilots, pop them in the cockpit, you don't have to do simulator training.

S1
Speaker 1
04:35
So there's a real age and advantage for Boeing if they can do that. To do that they have to demonstrate the plane doesn't fly any differently. And that is a sort of a key thing that sets them on this path of when the plane turns out, does fly quite differently, they end up going down this software solution to resolve the problem. And it just, it cascades.

S1
Speaker 1
04:58
Over the years of development, they end up making decisions that gives essentially this automation more and more and more power and then ultimately makes it dependent on 1 single sensor. So what you really have if you want to understand the story of why a single sensor crashes a plane, you really do have to go back over that 10-15 year history to say what happens in a company that produces an environment that that's all it takes.

S2
Speaker 2
05:22
Well it's almost like a single operator dependent system is what you're really boiling down to when you start talking about this because you look at it and you go okay this 1 thing goes wrong everything's going haywire.

S1
Speaker 1
05:32
Absolutely.

S2
Speaker 2
05:33
So when you started to take a look into this and you started taking a dive into this whole aspect of Boeing, how many hours have you dedicated just to finding out all the information that you have? Because this was a 60 minute presentation, but the amount of stories that you're going into and the amount of data in detail I could only fathom the amount of stuff that you must have invested in being able to find out this information

S1
Speaker 1
05:56
Yeah, there's a fabulous book I can't remember the name of the author but it's called flying blind and it's brilliant and then there was 1 reporter particularly from the Seattle Times Fellow Irishman actually, I can't remember his name either.

S2
Speaker 2
06:07
But if

S1
Speaker 1
06:08
you look up Seattle Times, Boeing, he did incredible. I think he got the Pulitzer Prize for incredible journalism on what was happening inside Boeing. And I mean, that's still an unfolding story today.

S1
Speaker 1
06:20
But certainly, Seattle times are incredible in terms of their information So yeah a huge amount of time, but I tend to get intrigued by these these failures Particularly the ones where people go there's nothing to see here. You know it was a simple mistake It was operator error and you go no. I wonder how how did we get to here? So that was sort of what drove that fascination.

S2
Speaker 2
06:41
So when you start going through this and you're going around and you're presenting this, Has Boeing ever contacted you in regards of this presentation? Because this would be something where I would imagine most companies are a little bit hesitant of you're sharing this information, even though it's publicly available based on how you're able to find it.

S1
Speaker 1
06:55
Yeah, it's absolutely publicly available. I did have 1 Boeing employee come up to me after 1 presentation and he

S2
Speaker 2
07:01
smiled and he said, well, thanks for making me the most hated guy in the room. Oh, can I ask roughly where you were at in the world when this happened?

S1
Speaker 1
07:11
I was in Australia. Oh,

S2
Speaker 2
07:13
very nice, very nice. So As you're here and you're going through all of this stuff and you're going through this information and you go into these details of talking about salt and talking about peaks and valleys essentially is what you're talking about on how everything's are laying down and how these systems are able to do these things in fluidity and how you need to be paying attention to some of the things that you're not paying attention to. When you look at this, how are you looking at these different companies, these different organizations, what are they not looking at that they should be looking at?

S1
Speaker 1
07:46
Yeah, so it's incredible. And this is the big challenge in safety, which you know, you'll hear almost everyone who's involved in safety talking about this. But the problem is all companies tend to focus on the visual actual things that are in front of them, which tend to be the smaller injuries that are occurring again and again and again.

S1
Speaker 1
08:06
Now when it comes to the big failures and the things that cause the big failures, what we know essentially regardless of industry is that it's controls we thought were in place to manage these big hazards that are ineffective on the day. We saw this in a fatality study we did in Australia. Yes you have bad luck, yes you have human error, but what causes that bad luck and human error to progress to the point where someone gets killed is that the controls that should have been there to prevent that weren't working. Most companies seem to put in controls and then just assume they will work.

S1
Speaker 1
08:43
They check them, but they seem to be genuinely surprised when they don't work. The organizations who are really good at this stuff are the ones who have what's called a sense of chronic unease. They don't trust the green reporting on their controls. They go and find out where they're weak and they try and deal with them.

S1
Speaker 1
09:00
That takes a huge amount of resources, focus. It takes a company who doesn't mind bad news, in fact encourages bad news to flow up through it so things can be dealt with. And that's something that's not natural for us humans. You have to put in the practices in your organization to actually make that happen.

S1
Speaker 1
09:18
So I think that's partly the reason why people don't see these things. I say to companies, particularly when it comes to single fatality risk, it's not hazards that kill people, it's ineffective controls. That's what kills people. So focus on your controls and make sure they're practical because they probably aren't near as practical as you think

S2
Speaker 2
09:41
they are. So when you go into an organization and you start talking about them having ineffective controls, how receptive are they? Because I mean, you just walked in and said, hey, your baby's ugly, is essentially what you're telling them.

S2
Speaker 2
09:51
That's pretty much what you're telling them. Well, we tend to be direct speakers, so you don't tend to get

S1
Speaker 1
09:56
us in unless you're sort of prepared for what you're going to find. Yeah, it can be incredibly confronting for organizations because organizations tend to build systems and take comfort that the systems are there. They very rarely ask how effective are these systems at producing the outcome we want.

S1
Speaker 1
10:17
They very rarely close that loop on it. A really interesting thing to do, which a lot of companies don't do, is look at the checks on their critical controls and they'll all be coming up green. Look at what's actually hurting people. And when you look at those investigations, you find, well, controls had to fail for that person to be hurt, but those 2 data sources are very rarely put together.

S1
Speaker 1
10:40
And when they are, you just get this real richness of, well, we've clearly got a problem with how we check our controls, because we're still failing to incident on some of them. So yeah, you need a receptive company. We find that you end up, you need to be essentially working for the CEO or the board because if you have a non-receptive board or CEO, this stuff doesn't have a chance.

S2
Speaker 2
11:04
So let's play the other side there. So when the, have you interacted with any organization where the CEO is not very receptive to your ideas when you first come in?

S1
Speaker 1
11:13
Yeah, yeah, that happens from time to time, absolutely.

S2
Speaker 2
11:16
So when you go in there and you start saying, okay, this is where you're leaning, this is where you need to start going. What is the standard interaction there? Because some people have a very hard time, even when consultants come in, even though they're saying we need help, but we need help the way that we want help.

S1
Speaker 1
11:33
That's right. And it doesn't even the firm who really wants help and wants to be told the stuff, you go through this incredible process with them where almost everything you say turns into, but yeah, but you don't understand. So what actually happens is this, this, and this, and this, and they explain it all out to you.

S1
Speaker 1
11:56
And then what typically happens is we go and we say, well, okay, we'll test that hypothesis. We'll go to the next level and test the hypotheses. And you come back with their own data, which is always very powerful, and you show, well, this is not happening. So a really good example is you say, look, people aren't following these procedures because they don't work.

S1
Speaker 1
12:16
And then people go, yeah, yeah, but we've got a procedure for changing the procedures, right? So there's a way of doing it, no problem. So we'll go, sure, show us your data on how many procedures have been changed in the last year. And the answer will be none.

S1
Speaker 1
12:32
So you go right classic work is imagined work is done. You believe you've got a procedure that will capture any changes that need to happen but in reality nothing's happened. It's not been used. So that's we love that sort of data approach that you have to go and find the evidence.

S1
Speaker 1
12:48
And we do a 2 pronged sort of thing with clients. You go in and go, we'll have a look at your system, we'll have a look at your data, particularly your incident near miss data, because you can tell so much about a culture, about what's in that and what's not in that. And even the wording that's used in that. We did some work with 1 client who had a quite a serious fire and they said we've never had a fire before.

S1
Speaker 1
13:11
But when you get into the data they had loads of fires. But their people were writing small fire, small flame, glow, glow. And of course these were all precursor events that were just being lost. But then when you put that data in front of a client, they suddenly go, I've got a reporting problem.

S1
Speaker 1
13:28
And I say, yeah, because they're afraid to say fire in their instance. They're putting them in, but they're afraid to say what they really are. They're using that minimizing language. So do that, go talk to HR, find out why people get fired, put those 2 data sources together, then go out and talk to the people, and they'll tell you very quickly about, you know, the evidence you're looking at in the data, they'll tell you why that is the way it is.

S1
Speaker 1
13:52
And it's a really powerful combination in terms of trying to get to the heart of what's in an organization. Because I think we often find If you have an organization with a good system, it's always in the interactions between things that the problems are. And getting to grips with those interactions is the bit that people often miss. When you say, look, if you have a good system, but you're jammed over here, because this KPI over here is making these people behave in this way, and you've got a bottleneck.

S1
Speaker 1
14:24
And it's only in dealing with that KPI that you'll make that change. So that's, it's really, I love going in and going, how do you think your system works? How does it actually work?

S2
Speaker 2
14:34
So you're almost going in and telling them hypothesis tests verify through their whole processes to see how it goes across the board. Am I seeing that accurately?

S1
Speaker 1
14:43
Absolutely and sometimes if they're really open-minded they'll say to us, here are a set of beliefs we have in this organization, can you test them? And that's really powerful as well to go in and say, right well everyone believes this the way this happens, let's let's have a look to see if this is how it happens.

S2
Speaker 2
15:01
But a belief system and actually bringing science or validation of data is totally 2 different things. I mean, I can believe anything that I want all day long. It doesn't mean that that's exactly how our system is running.

S1
Speaker 1
15:12
Exactly. And that can be the most confronting. The most confronting can be to say, here's this belief you have as an organization in how you do something and that data doesn't support that at all. And that's not us saying that, that's your data saying that.

S1
Speaker 1
15:29
Because we find it's really, Everyone collects data on everything, absolutely everything. And then all they seem to do, most companies seem to do, is they look at the data that relates to KPIs. And all the rest of the data, they don't even look at. So a lot of the value we end up bringing is, we just read their data.

S1
Speaker 1
15:44
We do something they don't normally do.

S2
Speaker 2
15:45
You're giving away trade

S1
Speaker 1
15:46
secrets now. It is so simple but so important. But most people are too busy in the company to look at the data that doesn't pertain to something that they have to put on a graph to report at the end of the month.

S1
Speaker 1
15:58
And the value that's in that data, which you now can tap into with a lot of AI and that is really powerful to be able to pull that together and say, well, how did your organization really look when you look at the broader data set?

S2
Speaker 2
16:11
And so when you're bringing these things to light, when you're showing them information based on the data that they've collected. What is the initial shock? What is the initial response?

S2
Speaker 2
16:20
Because I'm sure that most of your conversations starting off of, our company's different, we're different than other companies, we do it better, but we are having these minute problems.

S1
Speaker 1
16:31
Yeah, so you definitely get that pushback exactly like you said and we talked about earlier. And it's very, we find it's very important to sort of do a two-stage thing. So go in and present.

S1
Speaker 1
16:43
Some people can actually get quite angry. They can, It's always emotional. It's never not emotional. And then you sort of have to stand back and sort of do another meeting when you come back to talk rationally about it.

S1
Speaker 1
16:57
And it's amazing, we'll do meetings where people will react very emotionally to it, be quite angry about it. You come back in a week and they'll sit there and go, I've looked at it again and again and again this week and I think these are right and then you start to work through to the solution. The mistake is believing it won't be emotional. Yeah.

S2
Speaker 2
17:24
So you're almost a psychologist, a human resource person at that particular portion when you start seeing that they're going through this emotional side of this. I was going to take a guess at here, correct?

S1
Speaker 1
17:37
Yeah, yeah, you mean you are... You're

S2
Speaker 2
17:38
a counselor. I mean, let's be realistic.

S1
Speaker 1
17:40
Yeah, because you're... You know, a lot of, particularly in high hazard industries where there's really big risks, the belief your system is not in as much control as you want it to be is frankly terrifying, particularly if you've been the 1 who's built it or are maintaining it and driving it and all that sort of stuff. So actually to be told, no you've got some quite significant vulnerabilities in this system or even worse, your people are gaming the system over here in this way to manage this and they're creating this risk.

S2
Speaker 2
18:14
Not our people, they don't game anything.

S1
Speaker 1
18:17
That's really hard to hear and you know it is the the difficulty in getting bad news to flow up and in a company is is is It takes some real strength to the top of that company as someone said to me once and I think it's brilliant boards can't cope with safety graphs that rise. It's as simple as that. And anything that makes the safety graph rise causes concern.

S1
Speaker 1
18:45
And the concern can get in the way of the rational debate about what it actually means. And that's the issue. I just say if we could just get boards to think about safety the way they do about finance, we'd make so many strides. You know, if we said to a company, right when you run your financial health of the company you're only allowed 1 KPI, what's it going to be?

S1
Speaker 1
19:06
Well they'd say well you can't do it. You'd say well you do that with safety, you know you just put the drifter on it and you have to go. You know when you're doing your yearly budgets you don't make up your budget and then check it at the end of the year to hope you were right, like we do in safety, you check it every month. If someone doesn't have a profitable month you don't fire them, you don't say to them try harder to be profitable, you say okay well what went wrong, what do we need to do differently, let's trial something, all right let's trial this, did it work, did it not, Let's not jump to conclusions too quickly.

S1
Speaker 1
19:34
Let's give it a couple of months. You know, it's really about trying to get boards to take that mentality, which they have and apply it to safety because safety is really no different in terms of we like to pretend it is, because it's quite terrifying that it's not, but it really is no different to that.

S2
Speaker 2
19:50
So let's talk about structure then. If you were going to structure an organization and you say, okay, this is where these different departments are going to go into, you already know I'm going to ask the safety question, because that's definitely where I'm leaning. Where would you see the safety department reporting to?

S2
Speaker 2
20:05
Would it be HR? Would it be legal? Would it be finance? Or would it be operations?

S2
Speaker 2
20:09
How would you lay out the structure?

S1
Speaker 1
20:11
It's got to go straight to the top. And that's what you see with organizations who have gone through major failures. You have to have a safety person with the year of the CEO because the chief financial officer is going to have the year of the CEO.

S1
Speaker 1
20:26
You need your safety person at that level as well and what you see with many organizations is that they don't have that person, they have a major failure, suddenly that person exists. That person has to be completely outside of financial stuff so that they can do their job without, you know, the monthly targets mattering or caring. And I think 1 of the mistakes, I've got this a little bit of a controversial theory of which I have no evidence whatsoever, but I think a lot of companies set themselves up, particularly for in the resources industry. We go, right, we want to get that out of the ground, that resource, whatever it is.

S1
Speaker 1
21:02
Let's design the company to get that out of the ground as efficiently as we can. And we design the company and we start going down that path and we lock that in. And then at some point we go, oh hang on, we're going to need a safety department over there on the side and we're going to need an environmental department over there on the side to police this system we've already built. But that system was never built to be safety first.

S1
Speaker 1
21:24
That system was built to be efficiency first, which means your safety people and your environmental people will always be seen as the police, the piece of grit that gets caught in the wheel and slows things down. So it's critically important that they're able to get to the CEO and say, this is what's actually happening. This is what my people are telling me is actually happening on our sites. Because if the person at the top making the decisions doesn't have that information, how would we expect them to do anything differently?

S1
Speaker 1
21:55
And in the big failures, what you always find is that the CEO and the board, they're not getting the information that exists usually that would have them make better decisions.

S2
Speaker 2
22:07
So, I've asked this question several times, of course, to other people in regards of how they look at this. And some arguments that have come up in the past is, I had to earn my spot on the board Why should someone that's coming from a safety department that doesn't understand our industry because that's normally kind of where it falls into

S1
Speaker 1
22:23
Yep,

S2
Speaker 2
22:24
why should they not have to earn the same way that I did?

S1
Speaker 1
22:28
Yeah, I mean this

S2
Speaker 2
22:29
is this We're just talking theory. I call it totes, theory of everything.

S1
Speaker 1
22:34
Yeah, I mean, the question is, if you were a board, why would you not want someone like that on the board? You know, we put people on boards with really good financial ability and all those things, and that's absolutely necessary, but we do have to get that technical balance because 1 of the big problems with safety we know is risk imagination, it's risk competence, particularly when it comes to big process safety events which by default catastrophic or significant but really rare And you need people who will be able to sit on that board and say to everyone else, I know this appears really rare, but this is the nature of these risks. And this is how we need to think about managing it.

S1
Speaker 1
23:15
So for me, it's a case of going, if you're a board, you need that expertise there. You're more vulnerable if you don't have it. So the fact you're in a high hazard industry, in my view, would be earning the position for someone to be on that board with that right expertise who can say hang on, let's just look at this from a technical perspective. In the same way as financial people will say hang on, this is risky financially, and people will listen.

S1
Speaker 1
23:40
It's that balance that's important. And it's when that balance gets out of whack that I think we really are at risk.

S2
Speaker 2
23:47
I think that the way that you have that set up is perfect because that's the way that it should roll out. That is the way that people should be looking at it, not so much of a earn your spot. And I say earn your spot, and I know that sounds terrible, but it's having that expertise right there.

S1
Speaker 1
24:01
Yeah, I think.

S2
Speaker 2
24:03
Okay, so Sean, let me ask a strange question. What do you have coming up next? What is there anything event wise that you're going to be doing that's open to the public that they can go in and see it?

S1
Speaker 1
24:12
Yeah, there's a 2 big conferences industry coming up in the next 6 months. They're both mining industry conferences, 1 in Queensland, 1 in New South Wales. I'll be presenting at both of those as well.

S1
Speaker 1
24:25
Some of the, well hopefully presenting at 1 of them. Abstracts are in. And some of the team hopefully will be there as well. And then we're doing another 1 down in New South Wales where we were talking about a lot of this sort of stuff as well.

S1
Speaker 1
24:36
I just love talking about organisational failure and why it happens. And I'm not, I don't sort of do the solutions to these problems. I'm much more interested in what causes them in the first place and are we trying to solve the right causes? Because it's certainly the organizational ones not necessarily the technical ones that we should be be throwing things out So that's the the big ones that are coming up for us in the next little while.

S2
Speaker 2
25:04
And then are you planning on doing more stuff North America side?

S1
Speaker 1
25:08
Not at the moment, but love to.

S2
Speaker 2
25:10
So in talking about North America, what do you think about here? Energy Safety Canada, what they've been doing in Banff. I mean you and I, we got put through a lecture last night by the CEO Murray which I thought was great telling us about the history of what they've been doing in Banff for the last 30 years.

S2
Speaker 2
25:24
It's a beautiful venue. I mean I think it's probably 1 of the best venues that I've been to.

S1
Speaker 1
25:28
It's incredible. The conference is incredible. Over 900 people here.

S1
Speaker 1
25:32
Perfect balance between time to chat, sessions, long sessions which have been really good so there's time to really get into some meaty topics. The location is incredible, it's snowed. I live in Australia, I haven't seen snow in over 20 years. That was tremendously exciting, as was for the other Australians here.

S1
Speaker 1
25:51
It's been a fantastic conference.

S2
Speaker 2
25:53
Now we rode up together and, like we were having the discussion in the car, the pictures do not do this place justice in regards to how it looks. It's very impressive once you pull up and it's not that the pictures are not impressive it's just once you get here it's it's mesmerized

S1
Speaker 1
26:07
and with the problem of the pictures is this sort of kind of unbelievable you go it can't actually look like that and then when you get there you go oh wow it looks exactly like that

S2
Speaker 2
26:16
So if people want to find out more about what you have going on, where can they

S1
Speaker 1
26:18
go?

S2
Speaker 2
26:19
Go to our website,

S1
Speaker 1
26:20
BradyHawood.com.au, LinkedIn, join us on LinkedIn, that'll be great. We do a number of podcasts, which I would point people in the direction of. Again, they're on our website page, Rethinking Safety, which is all about safety, obviously.

S1
Speaker 1
26:34
Brady Habeo podcast, we talk about big failures and what happens and simplifying complexity. We talk about complexity theory in all sort of forms and parts of life, not just safety. But if you're interested in these sort of ideas, you'd probably like that as well.

S2
Speaker 2
26:47
Okay. Well, Sean, I really do appreciate you coming on.

S1
Speaker 1
26:49
Thank you very much, too.

S3
Speaker 3
26:52
The views and opinions expressed on this podcast are those of the host and its guests and do not necessarily reflect the official policy or position of the company. Examples of analysis discussed within this podcast are only examples. They should not be utilized in the real world as the only solution available as they are based only on very limited and dated open source information.

S3
Speaker 3
27:12
Assumptions made within this analysis are not reflective of the position of the company. No part of this podcast may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, mechanical, electronic, recording, or otherwise, without prior written permission of the creator of the podcast, Jay Allen.Speaker 2

00:08

Okay, we are broadcasting live because that's what we've been doing this week. Of course, having all kinds of fun times as we are hanging out at Energy Safety Canada in Banff. Yes, you know, that lovely location.

S2
Speaker 2
00:21
So let's talk. Sean, directly off the stage, we have you in the chair in front of us, probably 1 of the best presentations I think I've seen in my whole life and I'm not saying that because you're really good.

S1
Speaker 1
00:30
Thank you very much,

S2
Speaker 2
00:31
Jim. So talk to me because, okay, you covered this very intriguing incident with Boeing. Well multiple incident was Boeing How did you come up with the concept and why was it so important to you?

S1
Speaker 1
00:43
I Was fascinated by Boeing because well if I step back a bit I look at a lot of failures, investigate failure, but I look at a lot of famous failures as well. And Boeing is 1 of the best examples of a failure where you have a front line incident, 2 plane crashes, But the seeds of that failure are sown way back both up the corporate ladder inside the organization and back in time. So it's a very time-dependent failure that the company made decisions that locked them on a path that not made failure inevitable but increased the probability of problems.

S1
Speaker 1
01:16
Now we see other organizations do the same and I think for me we spend a lot of time talking about safety, the frontline, empowering them, that's incredibly important but there's this other side to it which is further back up the line. A lot of the problems the frontline have to deal with are set by people who don't necessarily intend to say you know cause problems but what they end up doing is create an environment where those problems essentially emerge out of it so for me it is that type of fear that very you know clearly shows that path dependence from from management decisions.

S2
Speaker 2
01:51
So so if because there's going to be people that are listening to this that were not here. Could you do a brief overview of what your presentation was about? And then we'll kind of dive into some of the more detailed information if you don't mind.

S1
Speaker 1
02:04
Yes, so Boeing comes out with the 737 MAX and within 5 months there's 2 crashes and 346 people are killed. After the first crash it's very much blamed on human error as is done in many of these cases that it was the pilots and the maintenance crew of the airline. What would transpire is that it was actually a problem with the plane and what had happened was the plane had a software system on board that could take control of the plane at certain points in the flight.

S1
Speaker 1
02:35
It relied upon a single sensor, the sensor which is an angle of attack sensor on the front of the plane that measures the angle of the plane essentially as it's flying. If anything goes wrong with that sensor, then it feeds erroneous data to this MCAS software system. The MCAS software system can take over the plane and cause serious problems, as it did in these 2 situations with the 2 faulty sensors. A lot of the presentation is really talking about the history of what happened in Boeing to produce a situation where all it took was a single sensor failure to crash a plane.

S1
Speaker 1
03:06
Because you know, in aviation, we have layers and layers of redundancy. So how did we get to a point without that redundancy? And it really goes back to, you know, Boeing was an engineering first company from 1916, by 1989, 140,000 people working in it. The belief was you build great planes and you'll be financially successful.

S1
Speaker 1
03:26
They had a great line they used to have in Boeing which was, you know, we hire engineers and other people. Engineers were encouraged to demand what they needed on their planes to have them safer. So that was the culture but then in 1997 they merged with McDonnell Douglas, very different type of firm, cost-cutting, return on shareholder investment was everything and over the next 10 years that the whole company morphed into a company away from engineering to a company where the money was was was king essentially and cost cutting was was everything in returning shareholder investment was everything And what you see as part of that is that when Boeing were responding to the challenge of Airbus's new A320neo, they had to work faster. They had to work within this new culture they created.

S1
Speaker 1
04:11
They started to say well Let's not build a brand new plane. Let's take an existing 737. We'll call it the Max we'll put more fuel efficient engines on it, but most importantly is we'll be able to say to the airlines it flies exactly the same as the old 737. Really important, because if you can say that to an airline, you can basically take their existing 737 pilots, pop them in the cockpit, you don't have to do simulator training.

S1
Speaker 1
04:35
So there's a real age and advantage for Boeing if they can do that. To do that they have to demonstrate the plane doesn't fly any differently. And that is a sort of a key thing that sets them on this path of when the plane turns out, does fly quite differently, they end up going down this software solution to resolve the problem. And it just, it cascades.

S1
Speaker 1
04:58
Over the years of development, they end up making decisions that gives essentially this automation more and more and more power and then ultimately makes it dependent on 1 single sensor. So what you really have if you want to understand the story of why a single sensor crashes a plane, you really do have to go back over that 10-15 year history to say what happens in a company that produces an environment that that's all it takes.

S2
Speaker 2
05:22
Well it's almost like a single operator dependent system is what you're really boiling down to when you start talking about this because you look at it and you go okay this 1 thing goes wrong everything's going haywire.

S1
Speaker 1
05:32
Absolutely.

S2
Speaker 2
05:33
So when you started to take a look into this and you started taking a dive into this whole aspect of Boeing, how many hours have you dedicated just to finding out all the information that you have? Because this was a 60 minute presentation, but the amount of stories that you're going into and the amount of data in detail I could only fathom the amount of stuff that you must have invested in being able to find out this information

S1
Speaker 1
05:56
Yeah, there's a fabulous book I can't remember the name of the author but it's called flying blind and it's brilliant and then there was 1 reporter particularly from the Seattle Times Fellow Irishman actually, I can't remember his name either.

S2
Speaker 2
06:07
But if

S1
Speaker 1
06:08
you look up Seattle Times, Boeing, he did incredible. I think he got the Pulitzer Prize for incredible journalism on what was happening inside Boeing. And I mean, that's still an unfolding story today.

S1
Speaker 1
06:20
But certainly, Seattle times are incredible in terms of their information So yeah a huge amount of time, but I tend to get intrigued by these these failures Particularly the ones where people go there's nothing to see here. You know it was a simple mistake It was operator error and you go no. I wonder how how did we get to here? So that was sort of what drove that fascination.

S2
Speaker 2
06:41
So when you start going through this and you're going around and you're presenting this, Has Boeing ever contacted you in regards of this presentation? Because this would be something where I would imagine most companies are a little bit hesitant of you're sharing this information, even though it's publicly available based on how you're able to find it.

S1
Speaker 1
06:55
Yeah, it's absolutely publicly available. I did have 1 Boeing employee come up to me after 1 presentation and he

S2
Speaker 2
07:01
smiled and he said, well, thanks for making me the most hated guy in the room. Oh, can I ask roughly where you were at in the world when this happened?

S1
Speaker 1
07:11
I was in Australia. Oh,

S2
Speaker 2
07:13
very nice, very nice. So As you're here and you're going through all of this stuff and you're going through this information and you go into these details of talking about salt and talking about peaks and valleys essentially is what you're talking about on how everything's are laying down and how these systems are able to do these things in fluidity and how you need to be paying attention to some of the things that you're not paying attention to. When you look at this, how are you looking at these different companies, these different organizations, what are they not looking at that they should be looking at?

S1
Speaker 1
07:46
Yeah, so it's incredible. And this is the big challenge in safety, which you know, you'll hear almost everyone who's involved in safety talking about this. But the problem is all companies tend to focus on the visual actual things that are in front of them, which tend to be the smaller injuries that are occurring again and again and again.

S1
Speaker 1
08:06
Now when it comes to the big failures and the things that cause the big failures, what we know essentially regardless of industry is that it's controls we thought were in place to manage these big hazards that are ineffective on the day. We saw this in a fatality study we did in Australia. Yes you have bad luck, yes you have human error, but what causes that bad luck and human error to progress to the point where someone gets killed is that the controls that should have been there to prevent that weren't working. Most companies seem to put in controls and then just assume they will work.

S1
Speaker 1
08:43
They check them, but they seem to be genuinely surprised when they don't work. The organizations who are really good at this stuff are the ones who have what's called a sense of chronic unease. They don't trust the green reporting on their controls. They go and find out where they're weak and they try and deal with them.

S1
Speaker 1
09:00
That takes a huge amount of resources, focus. It takes a company who doesn't mind bad news, in fact encourages bad news to flow up through it so things can be dealt with. And that's something that's not natural for us humans. You have to put in the practices in your organization to actually make that happen.

S1
Speaker 1
09:18
So I think that's partly the reason why people don't see these things. I say to companies, particularly when it comes to single fatality risk, it's not hazards that kill people, it's ineffective controls. That's what kills people. So focus on your controls and make sure they're practical because they probably aren't near as practical as you think

S2
Speaker 2
09:41
they are. So when you go into an organization and you start talking about them having ineffective controls, how receptive are they? Because I mean, you just walked in and said, hey, your baby's ugly, is essentially what you're telling them.

S2
Speaker 2
09:51
That's pretty much what you're telling them. Well, we tend to be direct speakers, so you don't tend to get

S1
Speaker 1
09:56
us in unless you're sort of prepared for what you're going to find. Yeah, it can be incredibly confronting for organizations because organizations tend to build systems and take comfort that the systems are there. They very rarely ask how effective are these systems at producing the outcome we want.

S1
Speaker 1
10:17
They very rarely close that loop on it. A really interesting thing to do, which a lot of companies don't do, is look at the checks on their critical controls and they'll all be coming up green. Look at what's actually hurting people. And when you look at those investigations, you find, well, controls had to fail for that person to be hurt, but those 2 data sources are very rarely put together.

S1
Speaker 1
10:40
And when they are, you just get this real richness of, well, we've clearly got a problem with how we check our controls, because we're still failing to incident on some of them. So yeah, you need a receptive company. We find that you end up, you need to be essentially working for the CEO or the board because if you have a non-receptive board or CEO, this stuff doesn't have a chance.

S2
Speaker 2
11:04
So let's play the other side there. So when the, have you interacted with any organization where the CEO is not very receptive to your ideas when you first come in?

S1
Speaker 1
11:13
Yeah, yeah, that happens from time to time, absolutely.

S2
Speaker 2
11:16
So when you go in there and you start saying, okay, this is where you're leaning, this is where you need to start going. What is the standard interaction there? Because some people have a very hard time, even when consultants come in, even though they're saying we need help, but we need help the way that we want help.

S1
Speaker 1
11:33
That's right. And it doesn't even the firm who really wants help and wants to be told the stuff, you go through this incredible process with them where almost everything you say turns into, but yeah, but you don't understand. So what actually happens is this, this, and this, and this, and they explain it all out to you.

S1
Speaker 1
11:56
And then what typically happens is we go and we say, well, okay, we'll test that hypothesis. We'll go to the next level and test the hypotheses. And you come back with their own data, which is always very powerful, and you show, well, this is not happening. So a really good example is you say, look, people aren't following these procedures because they don't work.

S1
Speaker 1
12:16
And then people go, yeah, yeah, but we've got a procedure for changing the procedures, right? So there's a way of doing it, no problem. So we'll go, sure, show us your data on how many procedures have been changed in the last year. And the answer will be none.

S1
Speaker 1
12:32
So you go right classic work is imagined work is done. You believe you've got a procedure that will capture any changes that need to happen but in reality nothing's happened. It's not been used. So that's we love that sort of data approach that you have to go and find the evidence.

S1
Speaker 1
12:48
And we do a 2 pronged sort of thing with clients. You go in and go, we'll have a look at your system, we'll have a look at your data, particularly your incident near miss data, because you can tell so much about a culture, about what's in that and what's not in that. And even the wording that's used in that. We did some work with 1 client who had a quite a serious fire and they said we've never had a fire before.

S1
Speaker 1
13:11
But when you get into the data they had loads of fires. But their people were writing small fire, small flame, glow, glow. And of course these were all precursor events that were just being lost. But then when you put that data in front of a client, they suddenly go, I've got a reporting problem.

S1
Speaker 1
13:28
And I say, yeah, because they're afraid to say fire in their instance. They're putting them in, but they're afraid to say what they really are. They're using that minimizing language. So do that, go talk to HR, find out why people get fired, put those 2 data sources together, then go out and talk to the people, and they'll tell you very quickly about, you know, the evidence you're looking at in the data, they'll tell you why that is the way it is.

S1
Speaker 1
13:52
And it's a really powerful combination in terms of trying to get to the heart of what's in an organization. Because I think we often find If you have an organization with a good system, it's always in the interactions between things that the problems are. And getting to grips with those interactions is the bit that people often miss. When you say, look, if you have a good system, but you're jammed over here, because this KPI over here is making these people behave in this way, and you've got a bottleneck.

S1
Speaker 1
14:24
And it's only in dealing with that KPI that you'll make that change. So that's, it's really, I love going in and going, how do you think your system works? How does it actually work?

S2
Speaker 2
14:34
So you're almost going in and telling them hypothesis tests verify through their whole processes to see how it goes across the board. Am I seeing that accurately?

S1
Speaker 1
14:43
Absolutely and sometimes if they're really open-minded they'll say to us, here are a set of beliefs we have in this organization, can you test them? And that's really powerful as well to go in and say, right well everyone believes this the way this happens, let's let's have a look to see if this is how it happens.

S2
Speaker 2
15:01
But a belief system and actually bringing science or validation of data is totally 2 different things. I mean, I can believe anything that I want all day long. It doesn't mean that that's exactly how our system is running.

S1
Speaker 1
15:12
Exactly. And that can be the most confronting. The most confronting can be to say, here's this belief you have as an organization in how you do something and that data doesn't support that at all. And that's not us saying that, that's your data saying that.

S1
Speaker 1
15:29
Because we find it's really, Everyone collects data on everything, absolutely everything. And then all they seem to do, most companies seem to do, is they look at the data that relates to KPIs. And all the rest of the data, they don't even look at. So a lot of the value we end up bringing is, we just read their data.

S1
Speaker 1
15:44
We do something they don't normally do.

S2
Speaker 2
15:45
You're giving away trade

S1
Speaker 1
15:46
secrets now. It is so simple but so important. But most people are too busy in the company to look at the data that doesn't pertain to something that they have to put on a graph to report at the end of the month.

S1
Speaker 1
15:58
And the value that's in that data, which you now can tap into with a lot of AI and that is really powerful to be able to pull that together and say, well, how did your organization really look when you look at the broader data set?

S2
Speaker 2
16:11
And so when you're bringing these things to light, when you're showing them information based on the data that they've collected. What is the initial shock? What is the initial response?

S2
Speaker 2
16:20
Because I'm sure that most of your conversations starting off of, our company's different, we're different than other companies, we do it better, but we are having these minute problems.

S1
Speaker 1
16:31
Yeah, so you definitely get that pushback exactly like you said and we talked about earlier. And it's very, we find it's very important to sort of do a two-stage thing. So go in and present.

S1
Speaker 1
16:43
Some people can actually get quite angry. They can, It's always emotional. It's never not emotional. And then you sort of have to stand back and sort of do another meeting when you come back to talk rationally about it.

S1
Speaker 1
16:57
And it's amazing, we'll do meetings where people will react very emotionally to it, be quite angry about it. You come back in a week and they'll sit there and go, I've looked at it again and again and again this week and I think these are right and then you start to work through to the solution. The mistake is believing it won't be emotional. Yeah.

S2
Speaker 2
17:24
So you're almost a psychologist, a human resource person at that particular portion when you start seeing that they're going through this emotional side of this. I was going to take a guess at here, correct?

S1
Speaker 1
17:37
Yeah, yeah, you mean you are... You're

S2
Speaker 2
17:38
a counselor. I mean, let's be realistic.

S1
Speaker 1
17:40
Yeah, because you're... You know, a lot of, particularly in high hazard industries where there's really big risks, the belief your system is not in as much control as you want it to be is frankly terrifying, particularly if you've been the 1 who's built it or are maintaining it and driving it and all that sort of stuff. So actually to be told, no you've got some quite significant vulnerabilities in this system or even worse, your people are gaming the system over here in this way to manage this and they're creating this risk.

S2
Speaker 2
18:14
Not our people, they don't game anything.

S1
Speaker 1
18:17
That's really hard to hear and you know it is the the difficulty in getting bad news to flow up and in a company is is is It takes some real strength to the top of that company as someone said to me once and I think it's brilliant boards can't cope with safety graphs that rise. It's as simple as that. And anything that makes the safety graph rise causes concern.

S1
Speaker 1
18:45
And the concern can get in the way of the rational debate about what it actually means. And that's the issue. I just say if we could just get boards to think about safety the way they do about finance, we'd make so many strides. You know, if we said to a company, right when you run your financial health of the company you're only allowed 1 KPI, what's it going to be?

S1
Speaker 1
19:06
Well they'd say well you can't do it. You'd say well you do that with safety, you know you just put the drifter on it and you have to go. You know when you're doing your yearly budgets you don't make up your budget and then check it at the end of the year to hope you were right, like we do in safety, you check it every month. If someone doesn't have a profitable month you don't fire them, you don't say to them try harder to be profitable, you say okay well what went wrong, what do we need to do differently, let's trial something, all right let's trial this, did it work, did it not, Let's not jump to conclusions too quickly.

S1
Speaker 1
19:34
Let's give it a couple of months. You know, it's really about trying to get boards to take that mentality, which they have and apply it to safety because safety is really no different in terms of we like to pretend it is, because it's quite terrifying that it's not, but it really is no different to that.

S2
Speaker 2
19:50
So let's talk about structure then. If you were going to structure an organization and you say, okay, this is where these different departments are going to go into, you already know I'm going to ask the safety question, because that's definitely where I'm leaning. Where would you see the safety department reporting to?

S2
Speaker 2
20:05
Would it be HR? Would it be legal? Would it be finance? Or would it be operations?

S2
Speaker 2
20:09
How would you lay out the structure?

S1
Speaker 1
20:11
It's got to go straight to the top. And that's what you see with organizations who have gone through major failures. You have to have a safety person with the year of the CEO because the chief financial officer is going to have the year of the CEO.

S1
Speaker 1
20:26
You need your safety person at that level as well and what you see with many organizations is that they don't have that person, they have a major failure, suddenly that person exists. That person has to be completely outside of financial stuff so that they can do their job without, you know, the monthly targets mattering or caring. And I think 1 of the mistakes, I've got this a little bit of a controversial theory of which I have no evidence whatsoever, but I think a lot of companies set themselves up, particularly for in the resources industry. We go, right, we want to get that out of the ground, that resource, whatever it is.

S1
Speaker 1
21:02
Let's design the company to get that out of the ground as efficiently as we can. And we design the company and we start going down that path and we lock that in. And then at some point we go, oh hang on, we're going to need a safety department over there on the side and we're going to need an environmental department over there on the side to police this system we've already built. But that system was never built to be safety first.

S1
Speaker 1
21:24
That system was built to be efficiency first, which means your safety people and your environmental people will always be seen as the police, the piece of grit that gets caught in the wheel and slows things down. So it's critically important that they're able to get to the CEO and say, this is what's actually happening. This is what my people are telling me is actually happening on our sites. Because if the person at the top making the decisions doesn't have that information, how would we expect them to do anything differently?

S1
Speaker 1
21:55
And in the big failures, what you always find is that the CEO and the board, they're not getting the information that exists usually that would have them make better decisions.

S2
Speaker 2
22:07
So, I've asked this question several times, of course, to other people in regards of how they look at this. And some arguments that have come up in the past is, I had to earn my spot on the board Why should someone that's coming from a safety department that doesn't understand our industry because that's normally kind of where it falls into

S1
Speaker 1
22:23
Yep,

S2
Speaker 2
22:24
why should they not have to earn the same way that I did?

S1
Speaker 1
22:28
Yeah, I mean this

S2
Speaker 2
22:29
is this We're just talking theory. I call it totes, theory of everything.

S1
Speaker 1
22:34
Yeah, I mean, the question is, if you were a board, why would you not want someone like that on the board? You know, we put people on boards with really good financial ability and all those things, and that's absolutely necessary, but we do have to get that technical balance because 1 of the big problems with safety we know is risk imagination, it's risk competence, particularly when it comes to big process safety events which by default catastrophic or significant but really rare And you need people who will be able to sit on that board and say to everyone else, I know this appears really rare, but this is the nature of these risks. And this is how we need to think about managing it.

S1
Speaker 1
23:15
So for me, it's a case of going, if you're a board, you need that expertise there. You're more vulnerable if you don't have it. So the fact you're in a high hazard industry, in my view, would be earning the position for someone to be on that board with that right expertise who can say hang on, let's just look at this from a technical perspective. In the same way as financial people will say hang on, this is risky financially, and people will listen.

S1
Speaker 1
23:40
It's that balance that's important. And it's when that balance gets out of whack that I think we really are at risk.

S2
Speaker 2
23:47
I think that the way that you have that set up is perfect because that's the way that it should roll out. That is the way that people should be looking at it, not so much of a earn your spot. And I say earn your spot, and I know that sounds terrible, but it's having that expertise right there.

S1
Speaker 1
24:01
Yeah, I think.

S2
Speaker 2
24:03
Okay, so Sean, let me ask a strange question. What do you have coming up next? What is there anything event wise that you're going to be doing that's open to the public that they can go in and see it?

S1
Speaker 1
24:12
Yeah, there's a 2 big conferences industry coming up in the next 6 months. They're both mining industry conferences, 1 in Queensland, 1 in New South Wales. I'll be presenting at both of those as well.

S1
Speaker 1
24:25
Some of the, well hopefully presenting at 1 of them. Abstracts are in. And some of the team hopefully will be there as well. And then we're doing another 1 down in New South Wales where we were talking about a lot of this sort of stuff as well.

S1
Speaker 1
24:36
I just love talking about organisational failure and why it happens. And I'm not, I don't sort of do the solutions to these problems. I'm much more interested in what causes them in the first place and are we trying to solve the right causes? Because it's certainly the organizational ones not necessarily the technical ones that we should be be throwing things out So that's the the big ones that are coming up for us in the next little while.

S2
Speaker 2
25:04
And then are you planning on doing more stuff North America side?

S1
Speaker 1
25:08
Not at the moment, but love to.

S2
Speaker 2
25:10
So in talking about North America, what do you think about here? Energy Safety Canada, what they've been doing in Banff. I mean you and I, we got put through a lecture last night by the CEO Murray which I thought was great telling us about the history of what they've been doing in Banff for the last 30 years.

S2
Speaker 2
25:24
It's a beautiful venue. I mean I think it's probably 1 of the best venues that I've been to.

S1
Speaker 1
25:28
It's incredible. The conference is incredible. Over 900 people here.

S1
Speaker 1
25:32
Perfect balance between time to chat, sessions, long sessions which have been really good so there's time to really get into some meaty topics. The location is incredible, it's snowed. I live in Australia, I haven't seen snow in over 20 years. That was tremendously exciting, as was for the other Australians here.

S1
Speaker 1
25:51
It's been a fantastic conference.

S2
Speaker 2
25:53
Now we rode up together and, like we were having the discussion in the car, the pictures do not do this place justice in regards to how it looks. It's very impressive once you pull up and it's not that the pictures are not impressive it's just once you get here it's it's mesmerized

S1
Speaker 1
26:07
and with the problem of the pictures is this sort of kind of unbelievable you go it can't actually look like that and then when you get there you go oh wow it looks exactly like that

S2
Speaker 2
26:16
So if people want to find out more about what you have going on, where can they

S1
Speaker 1
26:18
go?

S2
Speaker 2
26:19
Go to our website,

S1
Speaker 1
26:20
BradyHawood.com.au, LinkedIn, join us on LinkedIn, that'll be great. We do a number of podcasts, which I would point people in the direction of. Again, they're on our website page, Rethinking Safety, which is all about safety, obviously.

S1
Speaker 1
26:34
Brady Habeo podcast, we talk about big failures and what happens and simplifying complexity. We talk about complexity theory in all sort of forms and parts of life, not just safety. But if you're interested in these sort of ideas, you'd probably like that as well.

S2
Speaker 2
26:47
Okay. Well, Sean, I really do appreciate you coming on.

S1
Speaker 1
26:49
Thank you very much, too.

S3
Speaker 3
26:52
The views and opinions expressed on this podcast are those of the host and its guests and do not necessarily reflect the official policy or position of the company. Examples of analysis discussed within this podcast are only examples. They should not be utilized in the real world as the only solution available as they are based only on very limited and dated open source information.

S3
Speaker 3
27:12
Assumptions made within this analysis are not reflective of the position of the company. No part of this podcast may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, mechanical, electronic, recording, or otherwise, without prior written permission of the creator of the podcast, Jay Allen.