Skip to content

Exploring How Resilient Cognitive Solutions Makes Tough Decisions Possible

Business as Usual

We all know that complexity is the enemy of decision-making! Today, we welcome Bill Elm, Founder of Resilient Cognitive Solutions, to detail how he and his company have dedicated themselves to simplifying complex decision environments. RCS has codified 35 years of decision support science into a methodology that cuts through complexity in the most challenging environments. Its team of certified cognitive systems engineers, software developers, human-factors specialists, and visual designers are masters at enabling advanced decisions. Bill's career began as a Cognitive Systems Engineer designing Advanced Control Rooms for commercial nuclear power plants, including an alarm management system that still defines the state of the art. He is one of the longest practicing Cognitive Systems Engineers, combining over 34 years of applied CSE experience in domains ranging from process control to national intelligence.

 

 

Transcription:

So good afternoon, everyone. Happy Monday, hope everyone had a wonderful holiday weekend, we actually took Friday off. But we have some people dial in because of their habit. And we had a chance to just have an informal conversation on Friday with a few people. But here we are today, it's business as usual. And it is going right into the springtime. And I'm pretty happy about the guests that we have coming up. And I'm joined today, as always by Jonathan kersting. He is vice president of all things media and marketing at the tech Council. And we actually were just talking a few moments ago to remind people that we do have a radio show as well as a magazine. And we're doing everything that we can to tell the stories of the amazing people across this region and beyond who are driving innovation. So I want to give before I introduce our guest, Bill Elm just bear with me for a few moments, because I want to give my deep appreciation to Huntington bank for being our partners all through this journey. And we launched this almost a year ago, a year ago, plus, and Huntington stepped right in, they are very passionate about the technology ecosystem and making sure that we remained connected during the pandemic. So if you don't know Huntington, get to know them, because they're really there to help provide support to you for a lot of things, not just banking, they do a lot of other things to make sure you get what you need. Also 40 by 80 that is the 501 c three affiliate of the Pittsburgh Technology Council. And that is where we lead many of our efforts to support entrepreneurship and workforce development in the region. Soon you'll hear more about some of the work that's coming out of there including apprenticeships. So the chat is on as always, and we've needed you. And if you have any questions, don't hesitate, Jonathan will keep his eyes on the chat. And on top of that, this is not an opportunity for you to sell your wares. or shine the light on yourselves. This is our time to shine the light on our guest mats, what we're focused on. So on that note, I am bringing to the forefront bill Elm. He is the CEO and founder of resilient cognitive solutions. And I want to welcome him here. Hi, Bill.

Hi, Audrey.

Good morning. Good afternoon. Thank you for joining us appreciate this on a Happy Monday. thrilled. Absolutely. So Lexi is going to put in the website of resilient cognitive solutions. So if you want to see anything about them, while we're talking as many people multitask, totally understand. So, Bill, before we talk about resilient cognitive solutions, I want to spend a few minutes getting to know you a little bit. And understanding your background, you've got an interesting background, you've been at this for a little while. And so where do you want to start? What do you think we need to know about you?

Oh, I so I have a bit of an eclectic sort of a background, which really also describes the company. So I've got a couple of degrees from Carnegie Mellon, master's of electrical engineering in 78. But then went back again in 86, for some PhD, artificial intelligence work. Graduated from CMU in 78, and went active duty military army. And that that whole experience sort of changed me in a big way. It reflects kind of in my challenge seeking for the for the rest of my career. What once, once you think jumping out of an airplane with 100 pounds of gear is hard, then you go to Ranger School, and find out what hard really is. And that that sort of changed me. A friend of mine just got elected to the Ranger Hall of Fame in his. In his speech, he commented that people, people need crucibles in their lives, people need life, life changing gates to pass through and, and that was one for me. So I believe I am the first first successful Ranger tab out of Carnegie Mellon University. So that's maybe another first thing on my on my list. So as I got off active duty and was looking for a challenge in the civilian side, I joined Westinghouse, right after Three Mile Island, where they were challenged to redesign control rooms, a complete, complete ground up reimagining of a control room to support the decision making under under obviously very stressful conditions. And the thing that appealed to me about that when I When I interviewed for the job, I said that that sounds really hard. How are you going to do that? And the manager at the time looked at me, so we don't know, we have to figure that out first. And that that challenge kind of got me and I've been hooked on this ever since. So I've been doing cognitive systems engineering since 1982, the field was actually named in 1983. So I've been at this for a while.

So talk about what that means, though. So hard. So the word hard is is a familiar word for you. And obviously, that's been everything that you've done. But But tell everyone what cognitive means.

And yeah, so. So one way to think of us is we take what people think of as human centered design and extended to be decision centered design, really using the same cognitive science that the AI community uses to build algorithms. We use it to design the entire environment, people experience. And the research shows that, that changing that that cognitive design, changing that decision center design, fundamentally changes decision making performance, in particular under pressure. So when the when the nuclear power plant is breaking in, in a novel way, or never, never designed way, never predicted way. How do you help people fight and win? How do people help people make decisions under data overload thrive in the big data environment they face today? at scale at speed? With a lot at stake?

So if so, it's do you actually do any work in terms of personality? In terms of stamina, and you know, in terms of pressure that's tied to that,

lots of the sort of physiological measured measures and things to sort of assess stress. But the real basis of the design is, is built on the research on what separates experts from novices. What what you're what you're really trying to get is near expert performance from everybody. So that what, why didn't novices make mistakes and how to experts not make mistakes and, and make more elegant decisions? How did Sullenberger Landa make it impossible landing in the Hudson with an airplane that that couldn't fly with a situation that sort of is not? Not not achievable, and yet he was successful.

So there are some helpful, um, you know, some helpful practices that many of us should be thinking about at sort of the macro high level before we talk about the company.

There are there there are a lot of In fact, we've just completed a small book on what we call brittleness, self diagnosis, where designers, developers, entrepreneurs can can sort of detect these common, these common pitfalls. We've got a list of about 28 of them, that are, ironically just repeated over and over. People, people step on these every time. And they're, they're very, very detectable and very fixable.

So can you Is there a link to that that little book that you have that maybe we can share with it on your website,

it is not yet on the website, it will be up soon. And we can offer that. It's actually some of some of our more recent government work is now being released for for public use.

So as you hire people, it must be really interesting in terms of how you hire people. And in terms of what you look for, besides their technical competence.

Yeah, what it's a multidisciplinary field. So we have a wide spectrum of backgrounds, everything from psychologists, to graphic designers, systems engineers, and a crew of what I call fearless software developers, that that will lean and build these these, these crazy ideas we come up with. But we really look for that I'll say it's a bit of that crucible effect carrying over, we're looking for somebody that when I describe one of these really hard problems, some people lean back and and look very put off by that. Other people kind of lean in and get a twinkle in their eye and say that that sounds like fun. That sounds hard. Let's do that. And that's we're looking for it. We call it the gleam. So we're looking for that gleam and that candidate that really wants to work on hard things and roll up their sleeves and make a difference.

That's, that's really interesting. So that's great. So you also have an intern program. Let's just talk about that while we're on the topic.

Yeah, we do. We've put a lot into that. And we bring interns very selective, very sophisticated selection process there. And we bring interns in every summer We also have a partnership with University of Pittsburgh in their Co Op program. And so we have recurring co ops come back to us through that. And we immerse the intern directly onto a project team. So their their business card says they're an engineer day one. And they're they're thrown right into the deep end of the pool. They work as a member of the team. And we in beside that we have a, I call it sort of a lateral training program where they get to see a little bit of everything we do. But they are responsible to that team lead, making customer deliverables. And often they get to make customer presentations while they're here. So we challenge our interns,

we throw them right in, you throw them right in. People, you find interns from all over outside of Pittsburgh or in in Pittsburgh, or

we recruit heavily in Pittsburgh, but we draw from all over the country. We've had, we've had interns from Chicago, Virginia, Florida, New York, it is always best candidate. And we adapted it in 2020 to go virtual, obviously with a pandemic. So we worked hard to try to recreate that same, that same immersive, deep end of the pool experience for the interns, and we had some returning interns, second year interns said that we'd achieved at least 80% of the in person experience so that I gave us pretty good marks for that.

And so just quickly because I see someone said Are they college students, obviously in the co op program at at Pitt, they are college students, right. But

that, you know, they are the they start with us out of Pitt the end of their sophomore year. But in general, they are college students. We've had some graduates, some non students that degreed folks that also wanted to be interns. In general, they are college students, either undergraduates, or even PhD students have come to us as interns. And that's how we get them for a couple of years sometimes.

When are you looking for what disciplines

there are a few programs in the country that actually have multidisciplinary programs called cognitive systems engineering. Other than that, we get them out of industrial engineering, we get them out of psychology departments, we get them out of human centered design programs like like CMU is HCI program. So it's we we cast a pretty broad net, because we're in such a multidisciplinary kind of a, an approach that we ping into lots of different departments.

So So we, we found out a little bit that you're attracted to hard, you're attracted to, right. You're attracted through punching above your weight and, and then you achieve it and then you continue to punching above your weight. So what made you start the company you want him to do something even harder?

Well, it since since starting cognitive, with cognitive systems engineering at Westinghouse, I left there and went to Carnegie group. So now there's CMU spin off with, you know, McDermott and Dennis Jablonski took it over, we were always sort of a piece of a larger organization. And then when Carnegie group went public and got sold, we were again a piece of a larger organization. And we spun out in 2007, to lose focus. So there's a there's a book that that that influenced me a lot, or at least aligned with my thinking, called the discipline of market leadership. And that sort of excelling at your niche was one of those disciplines. And that really describes us. So by spinning out, we can focus. I don't have to explain to some vice president what, uh, you know, what, cognitive systems engineering? Yes, it's what we do. So we built the place from the ground up to be good at this. The facility is tailored to this the training program, the intern program, the recruiting program, it's it is all aligned with what we do.

So you have spoken about the need for for companies to have what's called a cheap decision support officer. So what kinds of company should have that role? And do you ever provide that role for any interim period of time for any of the companies?

Yeah, we're having some fun with that one. Yeah, so that the, there's a lot of buzz and a lot of things written now about Chief Data officers and, and companies should be exploding and companies are exploiting big data, and big data analytics and data science. So they, they they feel sort of one of those crucible events as a as a chief data officer. But the real purpose when you read between the lines of the real purpose of that is to support operational decision making of a company. And that's all between the lines of successful data science and successful chief data officer. So what what we say, and what I really promote is that you almost need a pair, you need the chief data officer to do all that data curating that is absolutely essential to, to get off the ground and get the foundation in place. But you need a chief decision support officer to convert that into a form that those corporate executives and decision makers can and get that benefit and get that that value from that, that data investment.

So when I think of when I think of the term, this probably dates me is business intelligence, right? This is a little bit

Yes. It's, it's, it's along those lines, the and what we bring to the to the table is the ability to deliver that. So there's, there's, there's a lot of people that, try that and need that and acknowledge the need for that. And our decision centered approaches are aimed at exactly that. So so we're the how, and and we have you to your earlier question, for companies that sort of aren't aren't ready to take the leap and, and add a position called the chief decision support officer. It's, it's part of what we do. So it's sort of an implied service in in in, in something we can provide, to sort of translate that data into decision support.

So there's a quite there's, there's two questions out there, Jonathan,

why don't you just jump in? Up to Bill, thanks for joining us on the show today. It's great to see you again. Good to see you doing well. Above Bob capital, once new, have you hired any of the CO ops students after they graduated?

Actually, we have. So Cloud just started this, this January. So he graduated, and we brought him right on, really glad to have him. He's a he's a key member of the team already.

Fantastic. You know, what, if any impact does this type of decision making have on consensus decision?

We actually have a project going now with, with the the DNI research folks to add, I'll crudely call it crowdsourcing to add that sort of that that kind of wisdom of the wisdom of the masses input into an into a managed highly rigorous analysis environment, and that we're actually excited about that, and it's going very well. So we're still trying to adapt the decision making tradecraft to use that effectively, at again, at scale on hard problems.

Okay, so yeah, it's interesting now, right, with crowdsourcing, and all the different kinds of things going on and how people make decisions and scientific versus sentiment, yes, you're that you're living in that space as well, you probably have to teach, you probably have to teach a lot of leaders how to undo their thinking.

That that I get that question a lot. And what what tends to happen, what we find almost universally, is that all of all of the things that are ascribed as the sort of eccentric idiosyncratic behavior that the the boss developed over time, those are all coping strategies, because they've not had a good framework to actually formalize their decision support. When we offer one of these, these sort of models of expertise based systems. It's remarkable watching them sort of fall in and realign on that. We get we get comments from senior military officers all the time. along the lines of I've been waiting my whole career for this. There's there's, there's lots of good war stories behind quotes like that. And that's really a gut check for us that we're that we're on the right path when when you get that kind of a comment from an expert. That's, that's really, that really makes our day.

So I like to think of myself as a romantic capitalist. And I think you just described an example. I've been waiting for you.

Yes, exactly. Yeah, we do get some of it. Yeah.

So you have worked on your website, you've linked like some pretty big bold clients like Boeing, Raytheon, I saw Bachtel, Booz Allen, etc. Those are just a few. Are you able to talk about any of the projects that you've done for them? I know a lot of them you cannot

wait. Yeah, so there's several we can talk about and probably more than we actually have time for. One One recent one was for a big In supportive of one of those big companies for the army futures command, and they had made a significant investment in computer vision analytics, lots of lots of images being gathered massive data overload conditions. And they, they felt their computer vision analytics, were going to solve that problem. The system they put together, in fact, was rejected by the users, and was not being fielded, we got that we got that phone call late, late in the program, actually late in the late in the year. And we were left with a very short amount of time to try to reimagine that, that that entire adventure. And, and by applying our cognitive approach to to designing that larger environment, we call it the joint cognitive system of the human intelligence, intimately coupled with the technical power. The joint cognitive system, we we inverted the whole model to, to, into a system that did fit elegantly with those, those, those analysts that we're working with had data and totally changed that that around, it went from not being adopted to being a field and in fact expanded.

And so it's interesting because it sounds like and we spoke about this before the show, it sounds like you will go in and sort of help sometimes at the last minute, or sometimes where people are sending smoke signals to a project run for a long time. And you have to come in and sort of turn it around and put it on a new path.

Yeah, I call it a cathartic moment. Sometimes it takes a cathartic moment like that. Yeah.

And so the the readiness of the leadership has to be there. Right? I mean, the readiness of the leader, how do you navigate that because you're running a project. And then you also have to manage the executives, on the customer side? On the customer side?

Yeah. So in the case of a cathartic moment, that actually helps us that because that catharsis is felt at that leadership level, that that they need, they need help, they need something different. There, they are, then sort of more open to trying it a different way. Because to, to the established systems engineering community, this this cognitive centered design process is different. It has elements of that cognitive science in it that they they're not familiar with, and they don't typically use. So once they, once they get a big cathartic event, they're more open to listening to to the approach, and that the real answer to your question is by giving them early results, so we, we have customer engagements, literally from the 90 day mark, where they're seeing the concept designs are beginning to be sketched and rolled out and come to life. And that that's really the the gut check for them that that it's on the right course, and they can see the light. So that it's that early delivery of low fidelity prototyping and low, low fidelity design concepts early in the process that that makes them true believers.

And so are you. Are you hiring interns right now?

We are, we are where we're pretty much the end of the phase of the selection phase. So we have a, we have a couple of positions on committed right now. So yes, we are.

So let's talk about a little bit about the pandemic. Okay. Any Is there any use case for the government health officials during the Age of COVID? How could you calculate cognitive engineering helped? Kind of?

Well, so back to you asked earlier about sort of common pitfalls or things that people can learn from the the big, one of the big issues is evident in the the pandemic conditions the, the lag in the, in the data is a classically difficult thing for people to manage. Humans don't do well with those, those temporal offsets in their brain. So from the time a test is taken to the time the test results come back to the time the patient either exhibits system symptoms are not those things are classically difficult decisions for humans to to connect to the data and actually then make operational decisions. And there's a class of display that we named, quickened displays that, that use some some specific techniques, that as soon as you install those, they're, they're similar to what's used on the bridge of a supertanker, by the way for for similar reasons. Same idea lags in the system are hard to control. And as you incorporate those, it's a remarkable change. So, yes, the the Dr. Fauci should have been using a quick and display to to help him manage lags in his in his in his data set?

Well, I think Brian Kennedy on our team has a call into Dr. Fauci. So I'm hoping that we can show. So and we'll make sure that this advice is definitely passed on to him.

Oh, good.

So there's a question, Jonathan, there's another question.

Yeah. Great one here. What's new? Do you incorporate ethical frameworks into your system? If so, how do you decide which do you want?

Yeah, so there's a big, big discussion now is is about ethical AI. And just to the in two sentences, what our current approach is that that model of expertise establishes the framework for human ethics to be sort of overlaid on on top of that whole, that whole platform. So we currently don't embed an ethics engine per se, into these environments. We're leaving the ethics on the human side of the joint cognitive system.

So as we wrap up the show, you've shared with us a lot about yourself, you shared a lot about your, you know, sort of the crucibles and some of the you know, your your values based in there. What advice do you have for the rest of us who are out here listening to you? And talking about subject matter? That I think most people are very interested in? What what what advice do you have

related to decision support and cognitive systems? Yeah, so maybe maybe two things. One would be beware of the siren song of technology. It's very easy to get totally enamored by the latest, latest cool AI ml technique. We add one more layer to our neural net, and wonderful things will happen. If only my machine learning data set were a little better, wonderful things would happen. And it's really all about how do you build effective tools with the technology capabilities that are practical today. So assuming assuming AI replacement of a complete replacement of a human is just a bridge too far. So let's build cool stuff with the power we can. We can feel today.

Wow. Thank you. I want to thank you for joining us today, Bill.

Well, thanks for having me, is the

founder and CEO of resilient cognitive solutions. Go out to his website, you can find info about interns, you can find a little bit of info about the work that they do. And I think soon to be released is a little book that they've put together that might be filled with some helpful hints for all of us. So Bill, stay safe. Continue to soar. And thank you for being just another great company based here in Pittsburgh.

Thank you, Audrey. And and congratulations to the tech counseling in the pandemic as well.

thriving, Jonathan, this man on board What's up?

We have an awesome guest tomorrow. JOHN dick is the founder and CEO of civic sciences stopping by he's returned guests visit usually about a year. He's got some great insights as to how we're going to start clawing your way out of this pandemic. Great stuff.

So thank you. Thanks again, Bill. Thank you for your work. And thanks, everyone, for joining us here today. Enjoy your day and we'll see you tomorrow. And if there's anything we can do at the tech Council, don't hesitate to reach out. Thanks, everyone. You guys

Transcribed by https://otter.ai