Performing a Pre-Task Pre-Mortem: An Interview with Joe Estey

In this interview, we're catching up with our old friend Joe Estey to learn about what he calls a "pre-task pre-mortem." A pre-task pre-mortem is a discussion to talk about (1) the desired end-state of the task, (2) the process for completing the task, (3) the resources used for the task, (4) where failure is most likely to occur during the task, and (4) mitigations and controls necessary based no the pre-mortem discussion. Now, let's learn more about pre-mortems for safety and learning. The video of the recorded discussion is immediately below. If you're rather read, we also created a transcript, so just click the MORE button in that case.

Learning about Pre-Task Pre-Mortems

Let's start learning about pre-mortems with Joe Estey.

Introductions

Convergence Training: Hey there, everybody, and welcome. This is Jeff Dalto of Convergence Training, I'm a Senior Learning & Development Specialist, and we've got a really special guest for our podcast/webcast series here today. Today we have Joe Estey. Joe is the Principal Performance Improvement Specialist with Lucas Engineering Management Service. So we're really excited to have Joe, we think you're going to enjoy this, and with that let me say hi to Joe and welcome and, Joe, could you tell us a little bit about yourself please? Joe: Jeff, thanks, and thanks for inviting me in, I appreciate it. I'm basically a former chemical operator, first-line manager, and an operations person, who at a point in their career realized how important safety, and environmental compliance, and a safety-conscious work environment was to the operational world. And so I fell into safety by accident. So most of what I do, at one time was a corporation, where I did work with different teams, where I helped improve their safety culture, their processes, and taking a look at how they're doing things. And then that grew into a business in itself, and so I began doing that on my own in 1995 and with a training firm I established. Went on and did that for a while, and then I joined Lucas in 2016. Always with the same vocation; haven't really changed from investigating better practices, taking a look at what caused the accidents, and then helping people come up with and address corrective actions.

Introduction to Pre-Mortems and Similar Safety Discussions

Convergence Training: Cool. So Joe has agreed to come talk with us today about something he calls a pre-mortem, and I thought we'd start by defining that and triangulating exactly what that is, how it's similar to or different than some things other people may refer to as a before-action review (BAR) or some similar terms. So Joe, can you tell us what a pre-mortem is and a little about the constellation of similar safety techniques like that? Joe: Oh, great. A pre-mortem is done in two different areas. One would be at a higher level, like managing projects. Or it can be done at the task- or activity level, for managing the job. So, it really depends on the level of the activity. At the project level, it would be where a project manager or the project execution team gets together early in the project, and they predict where and when failure is going to occur. So if certain things aren't in place, and we go out and execute the task as we believe it's going to happen, where is the post-mortem going to occur? Unfortunately, we're going to have to learn more from the failure that we didn't prevent in the first place. So, at a project level, a pre-morten is all about failure from a project perspective. At the task level, it's about you and I going to a job site, discussing the scope of work, looking at the hazards, in addition to the hazards considering the mistakes and errors that are going to be made along the way, ensuring we have the right controls in place (or we may refer to them as "error reduction techniques"), and then once again saying "If this doesn't happen by 10 o'clock, this is our contingency plan; if we don't get our result by 2 o'clock; then this is our contingency plan. We want to do pre-mortems so we don't have to do post-mortems. Pre-mortems don't differ a whole lot from before-action reviews, pre-accident investigations, or pre-task planning. They're all about the same. The nuance with the pre-morten, though, is really an eye set on failure rather than talking about the steps of the job, and ensuring we have the right controls in place. Convergence Training: Alright. And if I can ask you two quick follow-ups on that. If there are two pre-mortems, project level and task level, are we going to talk about the task level today? Joe: Yeah, I think the task level is more universal and more applicable for people that may be listening to this, or that go to work every day and try to get the job done. Convergence Training: Great. And in your discussion of the task-level pre-mortem, you talked about you and I at a site. Who is participating in this pre-mortem--is it management and employee, or who? Joe: Oh, great question. It can take two phases, depending on what your work planning process is. So it could be the work-planning team: so you're going to go out to write a switching order, you're going to go out to dig up a tank and put it somewhere...or build even build a house. And so the pre-mortem during the planning phase is with the principal work and team, and you ask the same questions: where and when is failure most likely to occur? At the end of the day, where are we most likely to have failed during the day? So that would be the question at the work-planning team level. And then once the plan is written, or once we've got it in our head if the plan isn't written (because it's more of a skill-based activity), you and I are going to go out and ask that question again. And the reason is to ensure that the work as planned is the same as work as performed, and our start-work criteria is so well-thought out that we're not going to have to stop work along the way. So we'll have fairly firmly established start-work criteria, and that could be just the two of us saying if we're going to do a resource check and an equipment check, and we're going to remind ourselves what a de-energization process looks like, if we're working on a control panel or a pump, and we'll talk through that thing. So that becomes more of an activity-level pre-mortem. [optin-monster slug="zuzuvyij6mtvscihsoql"]

Identifying Hazards and Risks Before A Job Begins

Convergence Training: Alright, so thanks for that. Before we dive into the pre-mortem, one of the reasons I'm interested in this is I feel like this is a way for the organization to learn more. So my first question for you is, "What are some ways, at a broad, high level, that a safety manager can proactively work together with employees before performing a task to identify and anticipate hazards and risks and reduce those risks before the job begins?" Joe: Good one. The one thing you can do is to get involved in a dialogue about the errors and mistakes as well as the hazards. So, we do a pretty good job in a lot of work sites identifying physical hazards and industrial risk hazards, but not such a great job identifying some of the work-evolution errors, the things that are predictable after the event (we all knew it was going to happen, we just didn't want to think it was coming), and so the safety manager can ask a couple of key questions. One key question would be: in doing this task, what's going to be the most challenging thing we're going to face? I have learned from my own experience, it might helpful but it's not always welcome from the people you're talking to (especially at the higher-trained craft level), to talk about their mistakes. So I find working with linemen or heavy equipment operators, if they have done this job without incident for a while (or without consequence--they may have had incidents, it just hasn't caught up to them), looking at them and saying "OK, on this job, on the work we have to do today, where is the mistake most likely to happen?" usually doesn't endear yourself to that group. Number one, they haven't made that mistake, number two, they don't believe they can make that mistake, and number three, if they haven't made it, they won't know what it would look like, so how can they share it? So instead of asking them about mistakes during the pre-mortem, ask them about challenges, like: "You know, the work we're doing is interesting, and I'm curious about what the greatest challenge may be for you when you're doing that task." I have yet to have anybody not share a great deal of information when they're asked that question. People tend to share about challenges when they don't like to reveal or discuss their mistakes. So that's one, just asking along the way, "Hey, what are your greatest challenges?" It doesn't mean you're going to fail, it doesn't mean you're the one who's going to create them, but what are your greatest challenges? And number two is asking them "Have we ever done similar work to this in the past, and if so, what did that turn out like? Is there anything you learned from that event in the past that we can apply to this case?" And so the pre-mortems is one of those "lessons applied from the past to the current experience we're having now." And you ask them--was there any downside to the activity the last time we did it? And you just kind of walk them through that--but it has to be an active dialogue.

Pre-Mortems and JHAs

Convergence Training: Gotcha. So another quick intro question: how is this different than, or how do you use this in tandem with a job hazard analysis? Joe: That's a really good question. A job hazard analysis has a maturity that the pre-mortem doesn't yet just because of the limited time in service. And so, for years, people have been refining the task hazard analysis, job hazard analysis, craft hazard analysis...I've worked for a lot of organizations, and they have a lot of names for that. But it's usually the same:

• Scope of work, by step
• Attached to the hazard you're going to face
• Attached to the required or recommended control

So that's the job hazard analysis. But it's (the JHA) mainly looking and industrial and physical hazards--fall from, caught-in between, struck bys, etc. It's not looking at the mental modes, the mistakes, issues like rushing, complacency, mental acuity work where you might be asked to keep aware of 2-3 things on the job at the same time. Those are the things that you have to actively get people to discuss just like a JHA about the hazards, you have to get them to discuss them (with the pre-mortem) about their mistakes. So usually, they're best used in tandem. Say you're going to tie off on a mezzanine. And you're tying off because your fall-protection plan requires it for your company. So in addition to doing work at an elevated height, and understanding some of the industrial risks, what are some of the challenges you might face during that time, and what are some of the errors that have been made in the past that you might want to avoid? So you would add that question into it. Convergence Training: I guess based on what you're saying there, the pre-mortem is going to be a bigger, more systemic look, and it will help tease out what you mentioned earlier, the difference between work as imagined or planned and work as actually performed. Joe: Yes, that's exactly right. [optin-monster slug="tfb0ea1ze5dftije3pdk"]

An Overview of Pre-Mortems

Convergence Training: OK, great, so the pre-mortem. Let's talk about it. Before we dive into the individual steps and talk about each one at a granular level, can you just give us the bird's-eye view of the sequence? Joe: Sure. First of all, you always start with the scope of the work: what is the intended final result? Hey, so when we're done with the day's activity, what are we supposed to have accomplished? And that helps you understand if they understand where they're going. So you know, sometimes a mistake is made because people don't understand what the end result looks like. So the very first thing is, let's just generally agree on what we're trying to do and what the end state looks like. So next would be, "What the process in achieving it? What are the steps we're going to take to make sure we get to that end result?" Then, number three, you normally ask about the resources, including time and tools. So are there any special resources, any time considerations we have to keep in mind? Are there things to thing about, such as "If we don't get this step done before lunch, is it worth not starting until after lunch?" Because we understand how circadian rhythms and affect safety, and there are better times to do some things rather than other times. Is there a work/rest regimen that has to be built in, maybe because of heath or mental fatigue because you're doing a task that requires a lot of procedures and double-checking? So those are the kinds of things you talk about right up front:

• The end result
• The scope
• The resources

And then the next thing would be simply "At what point in time, based on what we discussed, will failure occur? Where's the most likely place that our plan becomes undone." Convergence Training: Alright. And are those the four steps? Joe: Yep, those are the four steps, and our fifth one becomes do we have a mitigation, defense, or control in place that helps soften the blow of the plan when it becomes undone, or that prevents it from happening? So do we have something to mitigate the consequence of failure.

Pre-Mortem Step 1: Scope and End State

Convergence Training: OK, great. So we've got five steps, and maybe now that we've got a roadmap, you can talk us through each step in a little more granular detail. Could you start with that first step about the scope and the end state? Joe: Sure. So the scope and the end state--typically, people have an idea of what needs to be done, usually based on what they've been asked to do. So somebody says "Hey, we've got a problem with our system. Can you guys go fix that?" That is a very vague, generalized statement. What are my limits? Am I able to turn valves or manipulate electrical equipment? If you want me to go out there and do work, you've got to be a little bit more definitive in what you want the end result to be. So, when I was a chemical operator, and we had a system problem, someone might come up and tell us "Hey, go on out there and T, R, and R the system so we can get it up and running again (which meant troubleshoot, repair, and replace). That is a wide margin of work orders--and by the way, all of them were acceptable until we made a mistake and they weren't. And so then it would be "I only wanted you to see what was wrong with the system, I didn't want you to take it apart." Well, you actually did, because when you said T, R, and R, that gave me a wide latitude. I was supposed to do anything, and you were only going to nail me if I got something wrong. So we realized back then that maybe something like "you need to replace the pump with a similar pump, and it needs to perform at this rate (60 gallons per minute, or whatever). OK, so now you're helping me. So failure can come from not knowing the end state. I work with a lot of crews that do decommission work, and so that means they go in and they decommission an area with beryllium, asbestos, radioactive contaminants, and nobody tells them what the desired clean de minimus looks like. So they'll say "Hey, we just need to get that place ready for the next step.' So what does that mean? How much effort? What would be clean in the regulator's eyes or what would be clean in your own eyes? You've got to get that one out of the way, and though it seems simple, it's not. You need to tell people what the desired end state is in clear and concise terms so that they know exactly what the desired state is when they get there. Convergence Training: I can see how that's important. I have a relatively non-dangerous and arguably non-exciting office job, but that's important in my job, too. Joe: Well, it is. If someone comes in and says "Hey, they're coming in for spring cleaning to vacuum the carpets tomorrow, you need to move all your equipment out," that sounds pretty easy, right. But what do you mean by "all" my equipment? Where do you want me to put it? What door are they going to be using? Am I going to put all my stuff in front of the door they're needing to use to get in here? It's easier said than done. Convergence Training: Or even something as boring as "What is this meaning about and what are we doing?" Joe: Yeah, exactly. That's exactly right, yeah. What's the agenda-what's the end state? [optin-monster slug="wqhlhsrtzmkopgekqqkp"]

Pre-Mortem Step 2: Process and Critical Steps

Convergence Training: OK, great. Tell us about step 2, talking about the process of the critical steps. Joe: Yeah. In training, we use the video Nine Ways to Peel a Banana, because everyone believes what they've always done is the best way. It suits them; it's comfortable. So, whether it is taking off a flange, or doing an inspection, or using a checklist to check the fire extinguishers--if you were to watch people doing their work, somehow they do it a little different than the next person. In some cases, it doesn't matter, as long as they get to the desired end state. But in some other cases, it matters a lot. So you can peel a banana nine different ways. Three of those require a knife, which means it's more inherently dangerous because it now uses a tool. A few of them could damage the fruit, and if you don't want that to be a part of your quality process, you've got an issue. And then some of them aren't productive. You can spend all day peeling a banana, you can be the safest banana peeler in the world, and nobody will ever hire you again. So if you have nine different ways to peel a banana, imagine how many different ways there are to perform maintenance activities. And so you might have the desired end state, but nobody has really explained to you the best way to do the job around here based on our expectations--here's the only way you're allowed to do the job here. So you've got to make sure that is part of the pre-mortem discussion, because that will be fed into the pre-job plan. Convergence Training: And do you see that as a top-down command or as more of a collaborative discussion? Joe: It's got to be a collaborative discussion. For one, because of investment. If people are always prescribed every step they have to take, they have no interest in it. Then the onus is on the person planning the work, not the person performing the work, so what's in it for them? It's called skin in the game. If I have no skin in the game, if you're the one who told me how to change the alternator in the car--which bolt, at which time, and how to do it in sequence--hey, it's your job and not my job. But if you want to talk about it, and come up with better ways, and we come to a mutual agreement that that's the way to do it, then if we have to change it and get back together and talk about it, well then now I'm in a different place. Because we're both co-invested at that point. Convergence Training: Gotcha. Joe: David Marx in the book Whack-a-Mole said it this way, and I think it's a great way of saying it: "The organization and the individual both have singular duties [and there are only two]. The organization has the duty to establish the expectation for the worker, and the individual has the duty to follow those expectations." And so I think that's pretty clear. It's not the individual's role or responsibility to establish the expectation. But the organization would be wise to listen to the individual who has to work under them.

Pre-Mortem Step 3: Resources, Time, and Tools

Convergence Training: OK, so step 1 is the end state. Step 2 is the process and critical steps. Step 3 has to do with resources, time, and tools. Joe: Yeah, and with the pre-mortem, one of the questions would be--"Given the conditions that we're working under, with our desired end state and the way we need to do the work, what are the possible downsides of any of the resources that we have planned?" And I really like the "downsides" question. You know, we did a lot of work planning, people put a lot of good work into the planning process, but what's the downside? Now it could that the truck that has the material didn't show up until 10 o'clock, when it was supposed to be here at 8 o'clock. Well, what's that going to do to us? Or it could be that we have someone spelling the crew, because they have to take a break, so what's that going to do? You've got to cover some of those things, because where failure happens, are not in those hazardous areas we face where we're managing risks from the industrial perspective, but in the mental mode phase, it's like we didn't think about that. And so that's the downside question, given what we have to do, with the scope and process we're using, what's the downside? [optin-monster slug="zdlccricb3pdf6voanq8"]

Step 4 of Pre-Mortems: Where Are We Most Likely to Fail?

Convergence Training: OK. Failure. Where are the wheels going to fall off? Joe: Exactly. And having discussed those three questions we just asked--so it's the end of the day, and a good way to phrase it is "OK, so we just got the job done, and something didn't go the way we expected." What's the most likely way that things won't go as expected. Let's just talk about that. Now at first, when you're doing this with crews, they're going to look at you a little vacantly, because they're not used to those kind of questions. They haven't anticipated those. And there's no reason to, because they haven't been asked. But if you ask them, pretend it's the end of the day, we're calling it quits, it's 4:30, we're going to hit the showers or get ready to go, "where will we have failed?" It could be that we had to suspend the job at a certain time, and we didn't talk about what suspension will look like. So do we leave the equipment running? Did we have to control some damper valve, because of the air flow? And so once you prime that mental pump, people begin to talk about these things: "Oh, it will be when the system is de-energized, and then when we open something up, we find out that they didn't de-energize the control panel to it, they took the 480 off because that obviously powered the pump, but they didn't know there was a separate circuit that turned the lights on." And then you can ask "Oh, so what are we going to do about that?" And that's another way of talking about the defenses we have in place.

Step 5 of Pre-Mortems: Identifying Mitigations and Controls

Convergence Training: Great. And I guess the fifth and final phase is planning to mitigate risk and putting in controls, is that right? Joe: That's exactly right. So, go back and look at some of those pre-mortem points of interest--we call them triggers. OK, so if the truck doesn't show up, later, what's our work-around? And we get buy-in right there at the job site because you can imagine the two-part walk-through--the work planning team has to think about those things, but they may not always have the essential mitigation; they may not know because they're not the ones doing the work. So the supervisor and the welder are the ones planning the job. But now it's the welder and the operator doing the job, and they've got to come to an agreement that, should this happen, and it's something that we don't want to have happen, what are we going to do before it does happen--what's our mitigation or our control?

An Example of Using Pre-Mortems

Convergence Training: OK, so a simple five-step process. You do it before you perform a task. You've got:

1. Define the project and end state--what are we trying to do here?
2. The procedure or steps--and you mentioned that's a collaborative discussion
3. Resources, time, and tools
4. Where might we run into failure?
5. What to do to avoid that by putting mitigations and controls in place?

I wonder if you could share a story, maybe from your own personal experience, of using this process out in the field before a task and how it created some benefit? Joe: Yeah. I'll use one from a group I was working with about two weeks ago. They were getting ready to go in and put a new surface throughout the facility due to a series of contamination incidents they'd had--various contaminants that were industrial hygiene in nature and radiological in nature. And in the planning of the job, there was a very benign product that basically looked like water on the safety data sheet (SDS): everything was 0, 0, 0 when it came to risk and hazards. And so they thought, well this is just like painting a house, it's not a big deal. But as they walked through the job, one of them said, "You know, instead of handling this stuff, where we can't get all the crevices and nooks and crannies, we ought to fog this--you know, put an air atomizer and spray it." So they called the vendor, and the vendor says "Sure, you can do that." And what the vendor meant, of course, is you can do that in accordance with the product's labeling. And the product's labeling says that under normal ventilation, used outdoors, this substance has those benign qualities associated with risk. Now, you've got to read that small print on the product. And since they had it in their head that this wasn't much of a risk, and they were going to use it inside and turn it from a hand roll, which is what it was designed for, to an air-atomizing fixative, they didn't consider that when they were going to blow this stuff in the room, that the building's ventilation would take it suspended in air and take it through the HEPA filters. So as they got through their discussion, in the pre-mortem, they said "OK, so we are switching the way we use the product," which is part of step 2 about resources. The desired end state stays the same--we want to put a non-permeable finish that you can walk into and not wear any kind of mask in there. So question 1, the desired end state, was answered. Question 2, hey, that was changed. We're no longer rolling, we're spraying. Question 3, do we have the resources, do we have the knowledge,--well, you know, that does raise a problem, because when we spray the stuff in the room, it's just going to go over to the HEPA filter, it's not going to land on anything, so we have to cut the room's ventilation off. And, we can't even have passive ventilation, so we're going to have to tape and enclose inlets and outlets. OK, so then comes the downside question. If we do that, what's the biggest challenge we're going to face? Well, it's August, it's going to be 105 degrees outside, there will be no moving air inside, so it will be about 120 degrees inside, and that's OK, because we're going to fog this stuff into the room and nobody's going to be in there, but wait: how does that affect the cure rate? So this has to cure, it's 6 percent ammonia, how's that going to affect this? Oh, we better call the manufacturer. So there was the identification of an error that led to a hazard but it wasn't one that was considered through the job hazard analysis--that question may not come up. And then the next thing is, in addition to fogging stuff inside the room, we have now created a more confined work space (even if it doesn't meet the regulatory definition) that we had, so what's that going to do to us? So what's the downside to that? So if we fail, during the course of this work we're describing, what's it going to look like? Oh, well, we don't know what we don't know as a result of turning it into a mist, and we don't know what we don't know about the cure rate, and we haven't really predicted how turning off the active air flow is going to affect the work environment. Well, then, what is going to be our mitigation and control before we continue? And so what happened was, they were able to create stage times, you know, do not enter, start work criteria, the amount of [inaudible] that could be done, before they entered. So it led to a variety of work steps that they previously would not have considered.

Thoughts on Sharing Knowledge Learning in Pre-Mortems Throughout Organization

Convergence Training: That's a great example, thanks. So I can see how that this method, applied in this way, can lead to a lot of spontaneous, ad-hoc learning moments. And the people who are participating in it are learning. But I wonder if you have any tips for taking that learning in a very localized context, and capturing it, distributing it, and sharing it throughout the organization so that it becomes part of a bigger learning effort? Joe: You know, Jeff, that is a great question. And that's probably the biggest challenge I face with most organizations. And even the ones I probably haven't met, probably still face that challenge. And that is, sharing the learning experiences across the organization. There's a great report that came out in Europe not that long ago, in a brand-new book called How Did That Happen?, just released probably three weeks ago, and a great deal of the book covers why weak signals are not picked up on. Number one, because they don't have a consequence, so we don't feel the pain, and secondly, most organizations are not structured to learn from each other internally. I had never thought about that before. But if you look at the construct, organizationally, of most companies, they weren't built to learn from each other. They were built to do work in silos and compartments, and be rewarded for such. And so, if you do have an occasion like the one we just talked about with the fixative, how do you help somebody in the same company down the road or on the same