On this episode of the Pipeliners Podcast, Rob Day joins host Russel Treat to discuss AI-based risk washing and its implications for safety training. The conversation highlights how artificial intelligence can complete required online training and assessments on behalf of individuals, creating a significant new risk for pipeline operations. The episode emphasizes the crucial need for companies to shift their focus from simply verifying that training has been completed to ensuring that individuals can actually demonstrate their skills.
AI based Risk Washing in Training Delivery Show Notes, Links, and Insider Terms
- Rob Day is the Managing Director at Cognisense. He specializes in helping organizations in high-risk and highly-regulated industries administer online training and assessments, safeguarding their reputations by ensuring that all training is verified, compliant, and legally defensible. Connect with Rob on LinkedIn.
- AI-based risk washing: Using AI to take training or complete regulatory requirements on behalf of individuals, creating a false appearance of compliance without actual knowledge or skill acquisition.
- Pipelining: The industry of designing, building, operating, and maintaining pipelines for the transportation of natural gas, crude oil, and other products.
- Risk washing: The act of giving lip service to or performing a perfunctory activity related to a risk, allowing a business to “wash its hands of it” and appear to be addressing the problem without actually mitigating it.
- Compliance vs. safety: The distinction between following the minimum legal requirements (compliance) and actually achieving a safe outcome. While there is overlap, compliance does not guarantee safety.
- Regulatory due diligence: The proactive efforts a business takes to ensure it is meeting all relevant laws, standards, and regulations.
- OQ (The Operator Qualification Rule) refers to the 49 CFR Parts 192 and 195 requirements for pipeline operators to develop a qualification program to evaluate an individual’s ability to react to abnormal operating conditions (AOCs) that may occur while performing tasks.
- API (American Petroleum Institute) represents all segments of America’s oil and natural gas industry. Its nearly 600 members produce, process and distribute most of the nation’s energy. The industry supports millions of U.S. jobs and is backed by a growing grassroots movement of millions of Americans. API was formed in 1919 as a standards-setting organization. In its first 100 years, API has developed more than 800 standards to enhance operational and environmental safety, efficiency and sustainability.
- The CRM Rule (Control Room Management Rule as defined by 49 CFR Parts 192 and 195) introduced by PHMSA provides regulations and guidelines for control room managers to safely operate a pipeline. PHMSA’s pipeline safety regulations prescribe safety requirements for controllers, control rooms, and SCADA systems used to remotely monitor and control pipeline operations.
- PHMSA (Pipeline and Hazardous Materials Safety Administration) ensures the safe transportation of energy and hazardous materials.
- Control Room Management is regulated by PHMSA under 49 CFR Parts 192 and 195 for the transport of gas and hazardous liquid pipelines, respectively. PHMSA’s pipeline safety regulations prescribe safety requirements for controllers, control rooms, and SCADA systems used to remotely monitor and control pipeline operations.
- OSHA (Occupational Safety and Health Administration) is an agency in the United States Department of Labor. Their mission is to “assure safe and healthy working conditions for working men and women by setting and enforcing standards and by providing training, outreach, education, and assistance.”
- HAZCOM: An acronym for Hazard Communication, a standard that requires employers to provide information to employees about hazardous chemicals to which they may be exposed.
- Lockout/tagout: A safety procedure used to ensure that dangerous machinery is properly shut off and is not able to be started up again prior to the completion of maintenance or repair work.
- Cathodic protection: A technique used to control the corrosion of a metal surface by making it the cathode of an electrochemical cell.
- Pneumatic testing: A method of pipeline integrity testing that uses gas or air pressure to check for leaks and structural integrity.
- ANSI/ASSE standard (Z490): A standard for accepted practices in health, safety, and environmental training. It provides guidelines for ensuring the identity of trainees and preventing technological or AI-based circumvention of the training.
AI based Risk Washing in Training Delivery Full Episode Transcript
Russel Treat: Welcome to the “Pipeliners Podcast,” episode 402, sponsored by EnerSys Corporation, providers of POEMS, pipeline operations excellence management systems — operations and compliance software for the pipeline operator to address safety program management, control room management, and field operations. Find out more about POEMS at www.enersyscorp.com.
[background music]
Announcer: The Pipeliners Podcast, where professionals, bubba geeks, and industry insiders share their knowledge and experience about technology, projects, and pipeline operations. And now your host, Russel Treat.
Russel: Thanks for listening to the Pipeliners Podcast. I appreciate you taking the time, and to show the appreciation, we give away a customized YETI tumbler to one listener every episode. This week, our winner is Jeff Cremin with Enbridge. Congratulations, Jeff. Your YETI’s on its way. To learn how you can win this prize, stick around until the end of the episode.
This week, we speak to Rob Day with Cognisense about AI-based risk washing and training delivery. Now, guys, you need to listen to this episode. This is a relatively new thing related to how AI can actually take training on behalf of individuals and what some mitigation strategies are for that reality. This is a new risk for us in pipelining. Hey, Rob, welcome to the Pipeliners Podcast.
Rob Day: Thank you, sir.
Russel: I have asked you on to talk about risk washing, which, before you reached out to me, was a new subject, a new concept. Before we dive in, would you mind giving us a little bit of an introduction? Tell us a little bit about yourself, your background, what you do.
Rob: Sure. I started more of an operational capacity and then changed over the years as I’m getting older. What I mean by that as far as I had worked in refineries, I was a firefighter in the military, as far as worked for insurance companies, this type of thing, and hazmat response, all that type of thing. It was very much in the weeds or in the hazard. It was very much in there.
Then I actually started working more at a leadership as a risk advisor on different projects. Then I ended up working as an advisor for insurance companies. Eventually, law firms, I worked advising and changing regulations and standards, and that’s where I ended up drifting to.
Then, in the later years, ended up getting into technology, and that’s where I’m seeing things like, obviously, it’s exciting what’s happening now with AI. In the last eight years, I’ve been mainly focused on technology and that interface with regulatory due diligence and risk management for businesses.
Russel: It’s interesting you have a background in firefighting. What kind of firefighting did you do in the military?
Rob: The firefighting was full gamut, so hazmat, aircraft structural, and as far as literally…
Russel: Were you in the Air Force as a firefighter?
Rob: Air Force, yeah.
Russel: I was an Air Force civil engineer. I spent many nights sleeping in the firehouse so that the colonel wouldn’t get woke up in the middle of the night when there was a bell that rung. I got a whole appreciation for what firefighting is really about.
Rob: Oh, beautiful.
Russel: It’s about getting up in the middle of the night and running into the unknown.
Rob: I’m very good at sleeping two hours and enjoying that and having a good sleep. I still sleep very little to this day, and I credit it all to my fire department time.
Russel: [laughs]
Russel: That’s not what I brought you on to talk about, but it is an interesting background. I will tell you one quick story, and this is a cool tee-up for what we’re going to talk about today.
My first assignment in the Air Force was in Oklahoma City, which was an Air Force logistics base. We had all kinds of equipment. We were using heavy paint strippers, and they were rebuilding aircraft and X-raying metal and all kinds of hazardous chemicals and all kinds of stuff.
I was in a meeting, and the fire chief, which was a civil servant, and the assistant fire chief, which was a senior NCO, were sitting at a table. The colonel asked fire chief, says, “What’s the mission of the fire department here at Tinker Air Force Base?” He said, “To put out fires.” He said, “You’re fired.”
He looked at the chief. He said, “Chief, what’s the mission of firefighting here at Tinker Air Force Base?” He says, “To prevent fires.” He says, “You’re the new chief.”
[laughter]
Rob: Nice.
Russel: That’s really what we’re talking about because what our mission is in pipelining is to move the product, prevent incidents. Simply stated.
Rob: Yeah.
Russel: That’s cool. Your background’s awesome. I want to talk about first, what is risk washing? Can you define what that is, what that means?
Rob: If I took your story there and put another person there in front of the senior firefighter there, if the answer would have been the mission of the fire department is to make it look like we are trying to prevent fires [indecipherable] , that still would have been a…
That’s risk washing, as far as the idea that we say, “We recognize the risk, and we are going to give it almost a lip service or an activity associated. We’ll give it an allocation of funds or time. Thereby, we can wash our hands of it and say we’re done.” This sometimes is done consciously, but, I think, most of the time, in industry, is done unconsciously.
Russel: It’s interesting. I talk a lot about compliance versus safety. In the pipeline world, we have to follow a set of legal requirements around pipeline safety. Operators get inspected to ensure that they’re doing at least the minimum requirements. However, that’s distinct from actually achieving pipeline safety.
When I say it’s distinct, there’s a lot of overlap, but compliance is not safety. Compliance is a predicate to safety, oftentimes, but compliance is not safety. What you’re saying is, if I’m just doing the minimum I need to do to be compliant and not really focusing on does it actually have a safety benefit, that’s what you’re talking about.
In other words, I’m checking the boxes and making sure I’m getting through the inspections, but I’m not really focused on safety.
Rob: 100 percent. You will also find that the compliance, by not achieving the outcomes on safety, they’re also not able to actually demonstrate the due diligence or that to the regulator that they achieved compliance.
They can say, “We did a thing. It relates to the compliance requirement,” but because it wasn’t achieving any outcomes, it thereby not only didn’t manage the risk, but also the regulator would come back and say, “Your activity didn’t actually meet the requirements under legislation.” Most of the time, it’s performance-based legislation.
Russel: That’s true. We can get spun in that whole thing. I want to just say that I work with a lot of operators. All the operators I work with are very concerned about safety. If they end up in a risk washing thing, it’s not with intent.
Rob: Agree completely.
Russel: There’s a lot of things about human nature, and I call it gravity, where we just normally go, where gravity will take us to the place that we’re doing risk washing. There’s a level of deliberateness and analytical thought required to actually have it matter, have the outcome matter.
When we talk about risk in pipelining, there’s all kinds of elements. Some of those elements of risk are related to integrity management. More the thing that you work with is around operator qualification and how does training and execution of work relate to safety. Can you talk a little bit about this philosophically, how you guys think about that?
Rob: You bet. Actually, it may be good to think about…You talk about OQ and the other requirements and how it all relates into safety or actually managing that risk, if I put it that way. Anybody that’s doing pipeline work, there’s obviously the specific OQ requirements, the external ones, that everyone’s well aware of.
If we look at, as far as if you’re working inside a trench, now there’s also OSHA requirements in regards to trench safety. There could be confined space requirements because of the gases or contamination that could be there. Obviously, they get into line locating and gas detection. There is specific OQs, but then there’s also safety requirements that are coming out of that.
You end up utilizing training as far as as a way of managing all these expectations and requirements across the board. The OQ is obviously the ideal starting place, but if an organization only did OQ, that’s it. They didn’t cover HAZCOM. They didn’t cover how to respond to emergencies and anything like that. They would fall far short, both of the compliance side…
They would meet some of the OQ, but not the full OSHA/DoT…We can keep going, as far as all the extra requirements that are there.
Those requirements, whether it be regulatory or due diligence-related, often require training to be done. Most of that training has moved from in-person to online. That’s where I find the big gap has happened. When that move went from in-person to online, we lost a lot, as far as integrity.
Russel: There is a big difference between receiving content and demonstrating skill. There’s a big difference between demonstrating skill and understanding when things are abnormal.
Rob: 100 percent. I’ll actually add one more to that. There’s a big difference between the perception of someone receiving content and somebody actually receiving it. In the world of AI now…For example, I just finished a session where it showed that I did 10 hours worth of assessments. This was a high-stakes assessment. In 45 seconds, I got 100 percent using AI tools.
As far as OSHA 10, I can actually now, with the ability, get an agent to go in there and complete 10 hours of OSHA training. All I need to do is pay the funds to get the account in order to set it up, and it’s done. It’ll come out with saying that I completed X training, but I didn’t even…
The only human requirement with a lot of it now is a credit card, because, obviously, your AI doesn’t get access to that.
Russel: Let’s unpack that a little bit. I’m not sure everybody understands that. What you’re saying is pretty devastating, if you will. What you’re saying is that I can go on the Internet and buy access to an AI agent that will act as if it’s me and take the training.
Rob: Correct.
Russel: We’re not going to tell anybody how to actually do that.
Rob: Although, just so you know — I’m going to say this because everyone is going to quickly figure out — all you need to do is ask ChatGPT how to do it…
[crosstalk]
Rob: you start getting it. Or you look at the Reddit forums. That’s where a lot of it came up first, Reddit and other ones, where people are sharing how to do these things that are coming up. Literally, I tried it. This capability came out a couple months ago.
My learning curve was literally amount of five minutes it took me to go in there and said, “Go take this training.” I give it a URL and said, “Go there.” It came back and said, “I need your username and password.” It’s like, “Sorry. Forgot. Here’s my username and password.”
It came back and said, “It needed a credit card.” It went as far as it could, waited for my credit card. I put it in, hit send. Then I walked away. In 10 hours, I got an OSHA 10 certificate that came out the other end.
As you can imagine, this is concerning. We think of that job-specific training, like lockout/tagout, confined space, cathodic protection, pneumatic testing and everything, CPR. We can keep going. All of these things, most of them, are now done online and now can be requested.
If you gave me a system, just to underline how it’s like…This isn’t course by course. Say if I worked for X company. You had 15 pieces of training that required to do. You just give it the username and the password to log in. It will go through, navigate the menu system, and complete all of them.
It’s not like I have to even manually go and ask it to do each one. Just saying, “Here’s my company website. Please complete all training that is assigned to me.” I can even say, “Make sure I get 85 percent” or something. “I don’t want to be perfect.”
Russel: [laughs]
Russel: Take your time.” You can watch it as it’s doing it because, actually, what it’s doing is taking over your screen. You can see the mouse moving and everything else. You can tell it to be slow. You can tell it to watch the video, do the matching answer polls, answer security questions, the whole bit.
Russel: For pipeline operators and for those companies that provide online training for OQ, this is a big deal. It’s a big deal.
Rob: It is. We’ll say the reason I think it needs to be talked about directly to the operators is because in most cases, the operators…Well, I shouldn’t say, as far as…I don’t need…In the cases I’m aware of, that the operators are relying on third-party organizations to provide the training.
The interesting part, if you look at the regulatory requirement, it’s the employer that’s obligated to ensure the workers are trained. If they utilize a third party, that’s on the employer. The training organizations will be waiting for the employers to demand that they address this AI concern, but if the employers aren’t aware of it, they won’t request it.
The training providers, there’s no demand, obviously, to do so. They’re waiting for the employers. That’s why it’s important for this to be talked about at this time.
Russel: You’re absolutely right. In fact, we’re going to need to do a little special promotion of this episode so that we can get that word out. What are the mitigating activities that operators and third-party training providers ought to be taking in order to ensure people are really qualified?
Rob: The nice part when we’re talking to industries that we’re used to here, if you were to say, “What type of actions or mitigative actions need to be taken when you’re doing X project or job,” the first step is to do an assessment. Are we crossing a river? Are we going across a bridge? Is it underground? Whose land is it? What’s the topography? All these things add special risks.
Same as the process for any online training, online certification, qualifications or anything like that, is to go through and assess it. For example, I had somebody ask me…They say, “That’s OK.” What we do is we require everyone to turn our camera on, which is a big step. Then at least you know a human is there.
[crosstalk]
Rob: in a lot of cases. Depending on the risk, there is ways of…Obviously, we’ve seen the AIs spoofing the camera, as far as it creates an entity that’s there. Or, in this case — I’ll give you an example — the questions that you’re asking me, I could have an AI that actually creates that interface.
Almost like if you think of the President of the United States up talking. You always see, on those shows, where they have that glass. They’re looking through it. They’re reading the speech. It looks like they’re looking out at the crowd, but they’re actually just reading off the screen.
It could be doing that, where it would analyze the screen and everything you’re saying and be suggesting what I say. Right now, I could be looking and reading every word. Even an in-person process where we think we’re talking to the other person and they’re responding, or an interview, we have no way of knowing that that’s actually happening.
The case there was a fellow that just got a job offer from a large technology company. You would know the name. It’s one of the big five in the world. He got hired as a full-stack developer even though he had no development expertise.
When they asked him very technical questions on a video interview, this thing was going back and forth, analyzing what the question was, and giving him a response. He was just reading off the screen. It sounded like he had 10 years of development experience when he had none.
Russel: Oh, man. This stuff is scary, man.
Rob: It is. I would say, to answer your question — because you asked about where to go — and as far as assessment is the first thing, but the nice part is AI is giving as many tools out there to protect and manage this as it is circumventing it.
There is plenty of different solutions and ways of actually going and saying, how do we make sure that you are you, that you’re the one answering, that it’s not AI answering?
There is different tools there. I don’t want to suggest one because it pushes everyone in one direction and that only work for a small…
Are people doing it off the cell phone? Is it as simple as a site orientation where we’re letting what the alarms are going to sound like, where they’re supposed to go? For company, that’s one level. Are we talking about OQ qualifications or training and rigging?
Russel: I think, fundamentally, that we’re going to have to change our focus from making sure people receive training to making sure people can demonstrate skill.
Rob: Yeah, 100 percent.
Russel: That, fundamentally, is going to be the shift. You can still use the training, but if people are spoofing the training, they’re not going to be able to demonstrate skill.
Rob: Yep, 100 percent.
Russel: The way to deal with this is to focus on skill demonstration.
Rob: I would agree completely as far as that’s the end outcome. We still know that, as we have now, organizations are going to want the training to support. They ended up in court to say, “This is the things we did,” because maybe somebody has died, now we can’t prove that they had the knowledge.
Their actions certainly indicated or made it look like they didn’t have the appropriate knowledge. That due diligence requirement to want to prove it is going to be there.
I would say I like what you’re saying is, lett’s focus on can they do what they need to do, but when it comes to the training, just make sure that some basic assessments are being done.
I saw an organization that literally spent one day, added a couple of things, new technology and features that simply block the AI tools. It showed a level of due diligence that was leaps and bounds above what I see…
[crosstalk]
Rob: industry today.
Right now, give you an example, we have a [indecipherable] that has over 200 different qualifications right now, as far as training courses. Gosh, she got three demerits off her driver’s license. She has nuclear gauges, site training as far as site orientations, craning and rigging, forklift operation, you name it, lockout tag, gas detection, confined space, fall protection.
She has all of these things, and most of that can be addressed very simply. Can all of it be stopped? No. Not as far as even in person, there’s ways of getting by everything, but the goal is not to make it impossible, just not to make it the easiest path. Right now, the easiest path is just to enter a prompt and have it do it.
Russel: The reason we put locks on doors is to keep the honest people honest, right?
Rob: Yep.
Russel: If people are committed to cheating, they’re going to be able to cheat, and that’s not new. That’s not really the problem. I don’t know that that problem is endemic.
The other thing I wanted to do in this conversation is talk a little bit about the importance of not just delivering training, but verifying skill level. Do you have any examples in the pipeline space where risk washing, OQ, or safety training, implemented performance, or landed in a safety issue, anything like that, that you could talk about?
Rob: I’m trying to think of it. There’s different court cases and examples of organizations that end up getting fined. I don’t want to call any one organization out, but it’s not hard to find where they end up getting citation after citation, comes across the board.
This could be anything from the environmental sides, as far as how close they’re operating to rivers, where they’re getting rid of the spoil piles, this type of thing. It could be on the health and safety side.
The one thing to look at as far as anytime there is a risk in an organization, as far as how are they mitigating it, and in almost every case, they’re mitigating it through some sort of training as part of it, either directly or in support of whatever their mitigation is.
At that point, obviously, that’s when the risks happen. That’s the concern. I’m trying to think of how to do it without actually calling out any organization.
Russel: We don’t want to call names out. That’s certainly not what we do here. Maybe another thing we could talk about is the distinction between…Because your background, what you guys do is more OSHA-centric than OQ-centric. Maybe it’d be helpful to talk about what’s the difference between OSHA-centric training and OQ-centric training.
Rob: Sure. I’ll make it a little more broader. We actually work outside even OSHA. We work for medical organizations in the United States, Canada, UK, legal type of thing, as far as lawyers, accountants, finance, heavy industry, nuclear. We’re working across the board.
In your area — you’re right — you’re hit with both the OQ side of it and the actual OSHA side of it. We can even divide the OQ into the training that’s provided or whether it’s the actual assessment, the OQ assessment, at the end.
I would say, if we look at the OQ process, the nice part is, when you do an assessment that’s at a point, especially when you do it in person, you’re mitigating a lot of the issues. When you do it online, you’re creating a lot more without safeguards.
To answer your question as far as the difference between the OQ, the OQ helps with the base requirements or tasks specific to a certain aspect of the pipeline industry. Then the OSHA requirements are all the extras that come around it.
We talked about working in a trench or excavation type of thing. It’s a good example. It’s a natural hazard of pipelines. It’s there. As far as chemical exposure, we think of HAZCOM or HAZMAT or HAZWOPER and all these other things. These are all the extra OSHA requirements that are put on top and expectations that employers are putting in place to address and complement the OQ expectations.
Russel: The way I would frame the distinction is OQ is more how do you perform specific work. Like, I’m in a utility, I want to fuse a plastic pipe, that’s a very specific kind of thing. I need to know how to do that particular thing correctly. Versus the OSHA stuff is more how do I make sure that the environment in which I’m performing the work, that I’m safe to perform the work in that environment.
Rob: I can see that.
Russel: That’s more the distinction. That’s not 100 percent, but just generically, that’s not bad…
[crosstalk]
Rob: I like the way of putting that. You’re right. It’s never 100 percent because you get things like fall protection or…
[crosstalk]
Rob: showing them how to and making sure they can put on a harness and do things.
Russel: Again, it’s not always true, but generally, the OSHA stuff falls more on the construction side in pipelining or the pipeline facility side. The OQ stuff is more pipeline operations, opening/closing valves, launching/receiving pigs, that sort of thing.
Rob: Again — you’re right — with an overlap everywhere because even think of HAZCOM…
[crosstalk]
Russel: One of the most dangerous things you do in pipeline operations is launch or receive a pig. The most dangerous thing you do there is open the launcher or open the receiver. Part of that is because what’s in the pipe, the pressure and the actual material.
Rob: That hazard generated at the lockout/tagout, as far as controlling that energy before you open up and release the pig, is an OSHA expectation as well.
Russel: One of the things we talked about is some real-world examples. We talked about one. I’d like you to talk about this a little bit if you could. That is the Henkels & McCoy Pipeline Division incident from 2013. Can you tell us a little bit about that incident and what it was and what the contributing factors were related to OQ and OSHA training?
Rob: No problem. In that case, we’re talking about a gas pipeline contractor. They ended up getting fined. I believe it was around $70,000. They had multiple trench violations.
In this case, if I remember correctly, the workers were found to be operating in, I would think, about 11-foot deep trench without any kind of protective structures or anything in place to prevent a collapse. Obviously, these are OSHA requirements. In that case, there was multiple other citations that the organization had faced prior.
It almost exemplifies the requirements of the OSHA standards and how it relates to the OQ. There is the tasks, as you mentioned, but there’s also the environment that they need to control.
In this case, we’re talking about a trench, the risk of collapse on those individuals. They could be 100 percent compliant as far as the OQ requirements and still, in this case, were completely found negligent as far as the OSHA requirements. Both has that real-world impact on their operations.
Russel: If you were advising a pipeline operator, what are the things you would tell them to do? Let me do this. I’m going to do this a little different way. I’m going to talk a little bit about my takeaways from this conversation.
One is, if I’m doing any kind of online training, I need to make sure I’m doing something to ensure that a real person is taking the training, not some kind of AI or bot.
Rob: That’s a good way to put it. Actually, for these organizations, there is now an ANSI/ASSE standard. I shouldn’t say, “Is now.” It was revised. It’s been existed and used by OSHA for citations over the years.
That standard is saying, very similar to what you’re saying, we need to make sure we know who the individual is. We know their identity. We know they’re the one that participate in. We know AI didn’t do it or any kind of technological. You can say, “AI.” It can be vague. We just need to make sure that’s it.
If you’re doing training internally, make sure those things are satisfied. If you’re using external organizations, ask them whether they’ve achieved compliance with this new ANSI requirement. That’s health, safety, and environmental training.
Russel: Let me see here. Looking up something from your notes that you sent me. That’s ANSI Z490?
Rob: Correct.
Russel: We’ll link that up in the show notes so people can find that if they’re interested. The other thing I would say is my takeaway is that if I really want to get to safety performance, I need some way of doing skill verification in the field.
I would say that, ideally, that would become part of my inspection protocol for work performed in the field. I’d create some mechanism for capturing and measuring that over time so I can see performance over time.
Rob: I like that. If I were to be so bold as saying, if an organization says, “Where is the budget going to come to do that type of thing,” take it from online training that has no level of verification. If you don’t know that the right individual is doing it and it’s left this vagueness out there, that’s where it can go through.
Like you’re saying, can they actually do the job is more important. That’s where you mentioned, if you aim for safety and performance, you’re addressing compliance. No one is ever going to ask you about the compliance, usually, because nothing’s going wrong.
Instead of putting the money up towards the online training aspects where could it done by AI unverified, or, do it right. As far as online, there’s ways of doing…
[crosstalk]
Russel: This also ties to what a lot of people are doing in their pipeline safety management programs at present. They’re starting to look at, “How do we develop KPIs? What should those KPIs be? How do we track them,” all that kind of stuff. We’re very, very early, as an industry, in that process, but that’s where we’ve evolved to in the maturity of pipeline safety.
This conversation is going to get bigger because, really, the fact you delivered training is only an indicator of did you actually create skill and capability. It’s not real skill and capability. Real skill and capability actually occurs when the work’s performed.
Rob: Agreed. Actually, one thing you said there about the KPIs might be a good example of how an organization could end up inadvertently risk washing. If they start tracking what percentage of our population has done X training and their KPI is 99 percent has done it, they celebrate.
That doesn’t relate not only to the skills, that they develop the skills, but now we’re in a state where it doesn’t even relate to whether a human being even watched it or took it. Moving those KPIs to more, like you’re saying, performance-based and not patting yourself on the back because you have 99 percent compliance in a process that inherently…
[crosstalk]
Russel: That’s the other key takeaway, is that with AI and its capabilities, there are now risks that we didn’t have even five years ago.
Rob: 100 percent. I would even say even four months ago.
[crosstalk]
Russel: It’s moving quick, man. It’s moving really, really quick. It’s frightening because the capability in AI is moving quicker than we as people can understand the implications.
Rob: Mainly, it was because of my background…I mentioned law firms. I was involved with investigations where people were badly burnt and killed. I saw the importance of training. Then I ended up getting involved in the technology sector.
At this point in time, all I’m doing is helping organizations assess where are their weaknesses. We point and tell them where solutions are out there, but that’s the goal, just assess, at least understand your risks. As you know, the first step in managing risk is knowing that it’s present.
Russel: Then you got to build a program for doing it. The online training is not going to go away. It’s too cost-effective. It does fill an important need, but you got to look at this stuff holistically. Right?
Rob: Yeah. I would say one other thing. You mentioned the cost-effective and it’s not going away.
We had an organization that was actually called out. They ended up contacting us. They were called out mainly because they moved this very large project from in-person to online.
Obviously, we’ve talked about there’s some insecurities with or you lose some due diligence there by moving it online. They saved almost $8 million through the process. When the incident happened, obviously, that’s bad optics.
The idea is we moved it from an in-person training, where you could look at the person — I knew you were there, and I knew you were paying attention — to something that anybody could do or in this case, AI could do. The optics that came back to that organization were pretty devastating.
Russel: I get your point. I can also hear what the pipeline operator would say in response. That $8 million I’m saving is a hard cost out of my O&M budget. I can put it into other things that drive safety, where the incident is a soft cost until it actually happens.
Rob: Actually I wouldn’t blame them. I agree completely. I think the decision is, do we need the training? If we do, we find a way to do it in the most appropriate possible. We have so much legacy training. I haven’t met a contractor out there that said we have too little training.
Russel: That’s true.
Rob: There’s just so much training. It just keeps piling on, over and over. Do we need the training? Are we doing it in the most appropriate way? If not, maybe we can move those costs to someplace, like you’re saying, and verifying the results of training.
Russel: If we look at it as a program where part of the training’s online, part of the training’s in a classroom, and then we’re doing other things in the field to assess capability, then that works as a system and it’s not just about the training, right?
Rob: Beautiful, yeah.
Russel: You need to know what the effectiveness of your training is.
Rob: Well put.
Russel: Not just what you’re spending on it, but what the effectiveness of it is. You don’t know that unless you’re monitoring. You have a mechanism for monitoring, measuring skill.
Rob: That’s probably a good way to summarize, as far as, what is the response? Like risk washing. If you’re doing exactly what you said, then the organization is not risk washing. They’re not doing something for the optics of doing something.
Or even tricking themselves accidentally because of the fact that it was being done by others, or they always did it this way and they keep on moving that way without realizing they’re not getting that end result that they want.
Russel: There’s a lot of inherent inefficiency often in training, particularly around OSHA training.
Rob: 100 percent, yeah.
Russel: I recently did some work with a pipeline operator, and I had to get into — oh my gosh — probably eight different petrochemical facilities in the Houston Ship Channel area. There’s a huge training center down in Pasadena that does this.
I sat in that room for two days and I took some of the same training three and four times just because different operators required it. They had some little different twist on what they wanted, and they wanted you to check the box you got their training. Oh, gosh. It was excruciating. I understand the importance of it, but it was excruciating.
Rob: That’s my hope. When organizations start looking, assessing what they’re doing, one of the questions they’ll come up with is exactly, do we need to do this? Is there more efficient ways? Could we condense it, combine it?
Russel: Yeah, no doubt. Look, this has been great. I appreciate you coming on. I appreciate you guys reaching out to me and helping get this message out to pipeline operators that they’ve got a real risk.
I’m going to need to think a little bit about how we promote this because this is one of the ones that we need to elevate a bit, because I suspect most operators are not even thinking about this as a possibility right now.
Rob: I would agree because, again, as you mentioned, as far as the world’s changing so fast.
Russel: Yeah. No doubt, man.
Rob: It’s not anyone’s fault that they’re not aware of what’s happening.
Russel: No, just a what is, man. Just a what is. Look, thank you guys so much. Appreciate you, Rob. Hope to have you back.
Rob: Sounds perfect. Thank you, sir.
Russel: I hope you enjoyed this week’s episode of the Pipeliners Podcast and our conversation with Rob. Just a reminder before you go, you should register to win our customized Pipeliners Podcast YETI tumbler. Simply visit pipelinepodcastnetwork.com/win and enter yourself in the drawing.
If you’d like to support the podcast, best way to do that is to leave us a review. You can do that on Apple Podcast, Google Play, Spotify. Find instructions on our website at pipelinepodcastnetwork.com.
[background music]
Russel: If you have ideas, questions, or topics you’d be interested in, let me know, either by reaching out on the Contact Us page at pipelinepodcastnetwork.com or to me directly on LinkedIn. Thanks for listening. I’ll talk to you next week.
[music]