The Digital Landscape is Quickly Changing

00:00:00:04 – 00:00:01:04
Adam Pasch
Joe.

00:00:01:06 – 00:00:03:18
Adam Pasch
Oh, man, you went jacket!

00:00:03:19 – 00:00:08:18
Joe Monaco
No hoodie this time.

00:00:08:19 – 00:00:32:08
Adam Pasch
You’ve done a ton of thinking and a ton of production in automating workflows and injecting AI into processes where it’s beneficial. Tell me a little bit about some of the quickest wins people are overlooking.

00:00:32:09 – 00:00:55:13
Joe Monaco
Yeah. So, in terms of automating, augmenting, and accelerating an operator’s workflow, you know, the natural tendency is, “Let’s just rip out the legacy system and build it from scratch.” That’s often the last thing you want to do for a couple of reasons. One, that’s wicked hard when you look at all these systems of systems and all the data interfaces in and out.

00:00:55:14 – 00:01:21:22
Joe Monaco
So that’s part one of it. Part two is the change management aspect. The reason change needs to happen is because that operator, you know, has scar tissue. They live it every single day. And so the quick win is not replacing the legacy system. It is finding where the human tax is for that operator. What are they having to work around between system X and…

00:01:22:02 – 00:01:48:23
Joe Monaco
…system Y, for example? They have to swivel-chair from one network to another and, you know, open up one window and then another to copy and paste data. And so the quick win is really to figure out where the operator is, what his or her scar tissue is, and how you can layer technology across the stack.

00:01:49:01 – 00:02:04:04
Joe Monaco
I call it the stack, but really it’s the data layer, the application layer, and then the AI layer, in that order. So you figure out what the process gaps are, and then you fill them with technology, not the other way around.

00:02:04:05 – 00:02:37:10
Adam Pasch
Yeah. It’s so funny. This year, Brittany was just talking about this when describing working with both a data engineering team and an architecture team. Finding these opportunity points around these swivel-chair problems of system integration is, sadly, the human doing wildly manual, sometimes just data-entry tasks that are ripe with human error.

00:02:37:11 – 00:02:40:14
Adam Pasch
And also, who wants to do that job, right?

00:02:40:15 – 00:02:46:01
Joe Monaco
And they’ll always find a way. It’s just that it’ll be an eight-hour process instead of a four-hour process.

00:02:46:02 – 00:02:46:21
Adam Pasch
Exactly.

00:02:47:00 – 00:02:55:03
Joe Monaco
Which cuts down on bandwidth, velocity, and ultimately lethality when we’re trying to support government customers.

00:02:55:04 – 00:03:24:05
Adam Pasch
Absolutely. Well, so then, as we’re thinking about these overly manual processes today, injecting more automation where it makes sense, simplifying or maybe just cutting things out of the process where it makes sense, and then introducing AI or other tools, again, where it makes sense, what are some things that are overlooked within those specific options?

00:03:24:07 – 00:03:57:11
Joe Monaco
Okay. So I would say any time you automate, augment, or accelerate a human’s workflow, have the human in the loop. A big risk here is that you lose sight of data authority, provenance, and plausible deniability. So if, for example, you have a 13-step checklist and now you’re going to layer on an artificial intelligence capability to either automate that checklist or augment human behavior on that checklist…

00:03:57:13 – 00:04:27:17
Joe Monaco
At some point, who has ownership of step seven of 13 can get lost very easily. And that’s a risk. It’s not a reason not to do it. It’s just something we have to, you know, keep sight of as we innovate and digitally transform. And the way you do that is you cite your sources, you have an event ledger, and that’s how you layer on the AI layer.

00:04:27:18 – 00:04:50:19
Joe Monaco
Yeah, yeah. Because then you can go back and look. And when tempo surges, you’re going to have to have that credibility and link back to see why, for example, an airfield is suitable for airlift mission planning, or why a certain sensor works better in this environment than another sensor, before AI even touches it.

00:04:50:20 – 00:04:52:15
Joe Monaco
You need to have that foundation built.

00:04:52:16 – 00:05:25:12
Adam Pasch
Yeah. How do you go about figuring out which pieces of this, whether it’s a 13-step checklist or a 130-step checklist? How do you go about, I guess, piecing together the full solution that starts cutting it down from eight hours to seven, to four, to maybe 30 minutes? How are you going after it? Is it a bite of the elephant at a time, or is there a more holistic approach?

00:05:25:13 – 00:05:56:14
Joe Monaco
Well, it’s both, but part of it is connecting the system to the whole, which is the enterprise. Yeah. And so operational architecture is going to fill what we call the missing middle, which can take that mission intent, that commander’s intent, and those desired outcomes at the enterprise level and connect them to individual systems, which are part of systems of systems, and so on.

00:05:56:14 – 00:06:19:17
Joe Monaco
So there’s the whole aspect of it. There’s also the idea that where you sit is how you see the problem. Are you a data consumer or a data producer? Obviously, there’s classification. You have to work through data-sharing issues and all of those types of things. So we really have to work in that middle layer and look down and in while we’re looking up and out.

00:06:19:17 – 00:06:21:21
Joe Monaco
And you do that through operational architecture.

00:06:21:22 – 00:06:40:10
Adam Pasch
Yeah, right. So, as you’re bringing up operational architecture and this idea of the mission thread, I’d love for you to help me understand how I can envision what a mission thread is. And I think I’ve got a couple of follow-up questions. Okay.

00:06:40:11 – 00:07:09:07
Joe Monaco
Absolutely. So my definition is very military-biased. You have a commander, and he or she has an intent: “Let’s go take that hill,” or, “Let’s destroy this set of targets to shape for a follow-on operation.” So there’s some sort of concept of operations and intent that’s up at a higher level, right? Yeah. But then, in order to fulfill that purpose, you need a certain method.

00:07:09:07 – 00:07:39:04
Joe Monaco
And there are certain tasks. Some of them are very explicit, like you’ve got to take out a certain surface-to-air missile system. You have to do some sort of electromagnetic jamming. There are effects in space and cyberspace and other domains the Joint Force operates in, and then some are just more implied. And so, while we’re doing this, you know, we have to assume that the adversary-turned-enemy is not going to go this way.

00:07:39:09 – 00:08:03:03
Joe Monaco
They’re either not going to notice or they’re going to go that way. So mission engineering, and it’s very well defined in the Mission Engineering Guide and the Mission Architecture Style Guide put out by the Department of War, is about connecting those higher-level operations and that commander’s intent to the actual activities, systems, performers, and resources that have to fulfill that intent.

00:08:03:03 – 00:08:25:22
Joe Monaco
And there’s a missing middle in there, like we just talked about, where you have to really lay the foundation through operational architecture to know that, if I do this or change that down here at the system level, it’s going to have this effect up here at the enterprise level. And so that’s how I would explain that.

00:08:25:22 – 00:08:40:14
Joe Monaco
But it’s wicked hard to do, right? And you have to have the right capabilities, the right resources. You know, we have, what, 130 architects in our company? It takes a village to do this, right? It really does.

00:08:40:19 – 00:08:55:14
Adam Pasch
No, that’s super helpful. And then, first of all, with that being a mission thread, introduce for me a digital mission thread.

00:08:55:15 – 00:09:25:05
Joe Monaco
So, I’m not the right person to answer this, but I’ll just say, yeah, you can edit it to start later. But at some point, those CONOPS and the mission activities, you know, have to have a digital twin, or they have to be put into the right model-based systems engineering language to be able to connect things together.

00:09:25:05 – 00:09:29:16
Joe Monaco
And I’m not an engineer, an architect, or a developer.

00:09:29:17 – 00:09:30:11
Adam Pasch
But you play one on…

00:09:30:15 – 00:09:31:19
Joe Monaco
But I know a lot of people who…

00:09:31:19 – 00:09:32:04
Adam Pasch
Do it.

00:09:32:05 – 00:09:34:19
Joe Monaco
That’s right.

00:09:34:21 – 00:10:01:13
Adam Pasch
Well, so then, back to, I think, your mission owner hat: you’ve probably landed in a new organization every two to three years, and everything looks good for the first week. Then you start finding skeletons in the closet, the system skeletons in the closet. What are some of the scariest ones that you tend to find over and over again?

00:10:01:14 – 00:10:25:10
Joe Monaco
Well, probably the ones where, you know, two to three years prior to me getting to the new organization, they were having the conversation about what they were going to build, and there was going to be an increment of a software application. I even walked into an organization that had software fielded and in sustainment, and no operators were using the software.

00:10:25:10 – 00:10:57:11
Joe Monaco
So that happened. There are more overt skeletons like that out there, but an indicator of skeletons in the closet is when, you know, software development or engineering folks haven’t been talking to the actual operators. They’re working in a vacuum. They’re working in isolation through a program office or something. But there’s no iteration.

00:10:57:11 – 00:11:30:04
Joe Monaco
Because what I do know from a past life is that, to do it the right way, it’s a nonstop churn between staging, dev, and prod. You get capabilities in prod, and you’re still refining because you just don’t know. These are complex adaptive systems, right? There are emergent behaviors, things you didn’t anticipate. “Oh, now I need this data in here,” because now that it’s in production, I have other actors and performers interacting with the data.

00:11:30:05 – 00:11:59:05
Joe Monaco
Right. And we saw this a lot in my past job at the Air Operations Center because we started in the execution space. But there are so many things that happen in execution across airlift, air refueling, global air mobility, and aeromedical evacuation. And so, as you would put capabilities into production, a lot of effects were generated, some desired and some undesired.

00:11:59:05 – 00:12:08:14
Joe Monaco
But through an agile process that’s flexible, you could go ahead and fill those gaps pretty quickly. Fill the missing middle once the system hits the enterprise.

00:12:08:15 – 00:12:19:19
Adam Pasch
Yeah. So you’ve been out of uniform for half a year. What are some of the biggest things you’ve observed in that transition?

00:12:19:21 – 00:12:47:04
Joe Monaco
So, I actually wrote about this yesterday online. But the linkages between operational art and operational architecture have become very apparent to me, and we kind of just talked about some of that. You know, how do you link strategic purpose and aims to tactical actions? That would be the military’s operational art and how they plan, you know, at higher levels to execute missions.

00:12:47:08 – 00:13:15:02
Joe Monaco
On the operational architecture side, when we talk about system design and enterprise capability, it’s the same darn thing that’s happening, right? You have this missing middle in there, and you have a lot of different parts to the whole and a lot of different parts to other wholes: different program offices, different services, allies, and partners. All these things have to connect on a sunny day.

00:13:15:02 – 00:13:56:17
Joe Monaco
And then you go to a rainy-day scenario, and you’re contested in air, land, cyber, maritime, and space environments. Everything gets even more complex. And so planning on the military side rhymes a lot with operational architecture on the, we’ll call it, software development, cloud, AI, and cybersecurity side. And that was kind of eye-opening to me. But you have to have both, too, because if you’re over here filling the missing middle for operational architecture and you’re not keeping up with what the commander and the field forces need, then you’re building systems that don’t meet the need.

00:13:56:17 – 00:14:20:08
Adam Pasch
Well, right? And it sounds like it also means staying adaptable, so that you have this transparency from the National Security Strategy all the way through to execution. As strategy changes, as commander’s intent changes, and as other systems change, all of it continues to stay connected and effective.

00:14:20:10 – 00:14:39:19
Joe Monaco
That’s the ultimate mission thread right there: How can we give options all the way up to the National Command Authority when contested, as geopolitical conditions are changing dynamically and as the ways in which we’re being contested change, right, in all domains of operations?

00:14:39:22 – 00:14:41:22
Adam Pasch
Maintain an UNCOMN advantage.

00:14:42:03 – 00:14:50:12
Adam Pasch
I love it. Maintain, generate. It’s a systems thing, right? Continuous generation of that effect. Love it. All right. Cool. Thank you, sir. Thank you.