HSDF THE PODCAST

Part 2 of 2 - Future of Tactical Communications & Computing at the Edge

• Homeland Security & Defense Forum

Share episode

Welcome to “HSDF THE PODCAST,” a collection of policy discussions on government technology and homeland security brought to you by the Homeland Security and Defense Forum. 

In this second of a two-part series, hear how agencies like the TSA and CBP remain resilient while maintaining impeccable service continuity right where the action happens—at the tactical edge.

Featuring:

  • Craig Basham, Deputy Executive Director, IT Operations, Office of the Chief Information Officer, Department of Homeland Security
  • David Earle, Director, End Users Services Division, Transportation Security Administration
  • Charlie Armstrong, Chief Information Officer, Federal Emergency Management Agency
  • Jim Westdorp, Chief Technologist, Ciena Government Solutions, Inc. 
  • Dr. Dan Gerstein, Former Acting and Deputy Under Secretary, Science & Technology Directorate, DHS (moderator)

This discussion took place at the HSDF Technology Innovation in Government Symposium on February 22nd, 2024. 

Follow HSDF THE PODCAST and never miss latest insider talk on government technology, innovation, and security. Visit the HSDF YouTube channel to view hours of insightful policy discussion. For more information about the Homeland Security & Defense Forum (HSDF), visit hsdf.org.

Dr. Dan Gerstein:

A second, Jim David, tell us about TSA.

David Earle:

You know we are. I think we're in our early stages. I think it'd be tough to put a percentage maybe similar to FEMA, less than 10% there. We are working, you know, to get connectivity still working through our EIS transition to stabilize that connectivity at all airports. Looking at some RFIs with Starlink and the satellite communications at some tough to reach locations. You know we are working. Like I said, I deal with the hardware side of stuff a lot, so my bread and butter is getting that device out for those TSOs and everyone to use. We were recently just able to put out a couple thousand iPhones so then we can leverage some of that data that's being pushed out from all our systems. So that's the small step I see us taking. There's a lot more work to do on the edge computing but we're getting there.

Dr. Dan Gerstein:

Okay, jim, you wanted to come in on this question. Otherwise I'm going to start you off with another one, but go ahead.

Jim Westorp:

Yeah, just a little bit. You know, because you're asking. You know how things are going as we move towards the edge and I thought I could provide just a little perspective, because I support the DoD agencies, the civilian agencies and although I would say, you know, in terms of migrating increased capacity towards the edge, the tactical edge, if you will, that is just starting, but one of the things we're starting to see is, you know, a desire to leverage, as we get to the edge, more commercial-type technologies 5G being Class 1, lte before that and 5G now and many folks, of course, are going to be as a baseline, using the various carrier services. Yes, absolutely. But what do you do in places where there's not a carrier service? You've got a big base or you're out really at the edge. You know even some of the border areas where you don't have the kind of coverage you need. How do you turn up your own capabilities in that area? Even on some of the large enterprises inside buildings, they'd like to have an edge capability that allows them to use their handsets that perhaps they don't have. And even for some of the larger agencies, can they use particular parts of the spectrum that are allocated for government users? N79 happens to be one that we're doing some work in, so that you can use protected spectrum and get more secure services at the edge, and then still, when you walk back into Verizon or AT&T's coverage, you can still use that as well, and so we're starting to see a lot of interest in that and that's driving investments in equipment that supports a number of different protocols, that's able to do classical networking functions as well as the wireless functions.

Jim Westorp:

The other thing that computing at the edge is we're seeing more of that as allowing us to do is really to have more programmable and flexible edge elements. We're able to do things in software now that we couldn't before. Because we now have the ability. We have so much compute and storage available at the edge that we don't have to have fixed function hardware-based devices. We can make them software programmable, and so your capabilities can evolve as new software is available and as new missions. You can take the same equipment and program it to different things. So those are the kinds of things that we're seeing some of our large enterprise customers ask about, and it's just starting and we're starting to. I would say, yeah, 5%, 10%, but it's going to accelerate over the next few years.

Dr. Dan Gerstein:

Well, to follow on on that, because that's very interesting. I mean, are there any solutions that have been fielded, say across the government or even within industry, that you might recommend for the industry or for the government representatives that are here? I mean, are there exemplars that you think are worthy of them thinking about?

Jim Westorp:

Rather than throw some specific names out, which I have, but in this forum is a little challenging sometimes.

Jim Westorp:

But you know there was some discussions of, you know, edge to cloud and how do we get to cloud. There are a number of services, some of which Siena supplies directly, some of which other equipment providers provide, that make it much easier and much faster to get from where you are back to the cloud. So there's cloud interconnect solutions that are in fact available, specifically tailored edge wireless solutions, particularly in the 5G realm, that use protected spectrum, that are available for high security, enterprise class applications. So both terrestrial wireline and wireless solutions that are actually integrated with networks and can be brought together as a solution are available. One of the challenges that the industry has and that some of the consumers in the various large enterprises have is how to consume this in a way that's easy, because a lot of them aren't necessarily comms experts. So we in industry are spending a fair bit of time trying to make the solutions easier to access, to increase the level of automation, so you don't have to be a rank expert in order to take advantage of that.

Dr. Dan Gerstein:

Great. So we've talked a lot about the benefits that could possibly accrue from edge solutions. I just want to pick at this a little bit and ask our government reps here to talk a little bit about do you have any concerns as you're going down this path? You know, often you hear about cost being a concern. Sometimes you hear about security of data. Sometimes you hear about, you know, sometimes the data doesn't get evacuated back so you can lose your data if you're not having protocols that bring it back. So are there issues in particular that you're worried about and I'd like to start again, craig, if you'll lead us off- Thanks.

Craig Basham:

So two in general. One goes back to the security is always a concern, right, cyber security in our realm. The old days of having analog video or analog voice days and having radios and doing TACOM and not securing those, you know those days are kind of far gone. We've gone digital and so as we start pushing further and further out, we've got to start really considering the security around this, not only the device but the data that's coming back. So how are you protecting the core, especially when they start connecting into our big networks? How are you protecting the data at the edge and so we still really want to partner with the industry on this is understanding how do you protect that data?

Craig Basham:

The second is around the type of solution we put out. So right now, if you look at a 5G trying to push out that control and user plane further out, not every solution is going to fit that. So a lot of times you have put these big solutions in that has big get to the cloud, but you have all the control back at the central location. So how do you start pushing that further out to the edge, start pushing onto the user device and providing just the data you need to do the operations. So what we're seeing is a lot of times it's the well, you can take this same solution and put it everywhere across the country, and it just is not the same thing. You either have to have that architecture and want to be centralized and push it out, or you need to really look at different capabilities to do that. Okay.

Charlie Armstrong:

Charlie, I think Craig's point really complements what I was saying earlier, with us around, you know, really trying to get that cloud solution in place, because the points he's making are exactly the same kind of issues that we're dealing with in cloud, where, you know, we need to make sure we've got a common security environment wrapped around all of our CSPs that come together as our cloud environment. Cloud environment. We've got to have common access control so that, you know, I've got the same level of assurity that access to those edge systems is the same as it would be if it was a hardened system. You know, in a data center. Somewhere I have to worry about nation states somewhere. I have to worry about nation states trying to get to us, and we've already seen where they're really looking for areas around disasters to try to get in and get some intel out of. I think we saw that with Typhoon, volt and some other recent and some other recent things. I worry about those three things.

Charlie Armstrong:

I think the other big thing, though, is, as we look at SaaS solutions. I know that there's a lot of going to be conversation or was conversation today about AI, but making sure that those capabilities are designed in such a fashion that I can have pieces running at the edge that can couple back into the main core of the SaaS product, especially as we start to talk more and more about AI. I may need to have that small bot capability out on the edge, but at some points in time it may need to back end and in our environment, I could see where we're going to do probably a lot of analytics around things like weather and flood levels and all those kind of things before the disaster and that's information that we could pump down to those edge devices prior to the event happening. Right, so that we've already got some of that level of data out there. So, making sure that we're doing that pre-planning and knowing what is going to be needed in order to be able to accomplish the function.

David Earle:

David your thoughts. One of the benefits of going third with these guys is security as well, so I'll just echo what they said. You know security would be big for us. But I'll touch on another one Scalability. We have 600 plus locations around the country. Can I get that solution into? You know, I probably can get it into an lax, but can I get it into casper wyoming right? So it's, how do I leverage that solution to get out to all these locations, providing them the same data? Because they all need the same data.

Audience:

Uh, and how do I?

David Earle:

scale that, so I would see that as being one of our big concerns.

Audience:

Great.

Charlie Armstrong:

Can I add one more? Yeah, please, physical hardening. Physical hardening is going to be really key, especially in a lot of our lines of business because we operate in not typical environments and making sure that they can withstand weather, heat, all those kind of things is going to be really key.

Dr. Dan Gerstein:

Great. So we've got about 10 minutes left. I want to open it up for audience questions. But I want a very quick and it's a lightning round, and that is are there budgets? And I'll just, I just want just give me a thumbs up or thumbs down as they're getting, as they're getting their questions ready. You know, are you comfortable that there is money and I'm not looking for specifics here, understanding our budget turmoil, but more about conceptually. You know, do you have numbers in mind and are you working this through so very quickly, like a 10 second thumbs up, thumbs down, whatever?

Charlie Armstrong:

Senator, the administration has no official position on that. Thank you.

Dr. Dan Gerstein:

Okay.

Craig Basham:

CIO, how about you? Sure, there's money, there are always money. No, there are types of money for allocations for certain type of solutions. Good David.

David Earle:

Yeah, I'm similar to what Craig said. We have some avenues that we could pull in some funding to make some things work, okay.

Dr. Dan Gerstein:

So, questions.

Charlie Armstrong:

Dan, if I could add one more thing, yeah, please. I think this is a really key opportunity, though, for if you look at CBP and, I believe, tsa with these invent funds, I really think those types of things need to be thought about more across, not just DHS, but across the community. Allow us to try stuff out. You know, try quick things, feel fast or succeed quickly and move out.

Dr. Dan Gerstein:

I mean, I think that's a great point is being able to innovate, and I think the government is sort of changing its tone on how they feel about innovation and the need for a little more risk. Hopefully that change in tone will result in change in action as well. I saw a question here.

Audience:

Jason Cronin from Axiom Consultants. This is probably for Jim. It has to do with security. Obviously, security is an issue that's been mentioned a couple of times. But what do you feel is a long pole in the tent when it comes to endpoint or security at the endpoint? Yeah, I'm not asking what Sinus' solution is, but what your wrong poll on the tent is.

Jim Westorp:

Thank you, so, as far as endpoint security, there are a number of different solutions other than maybe getting the appropriate standardization in place so that when you're doing security in one place, you have the ability to do that in multiple other places. One of the things that we as an infrastructure company and industry writ large is working at doing is being able to secure the information at multiple layers of its transport. So, absolutely, you want to be able to secure it at the handset, but you've got threats at multiple points in the network, so you actually have to secure it in multiple places. Yes, at the endpoint, but also as you go deeper in the network, you want to protect those parts of the network as well. So you know, the problem set, or the long pole, if you will, I would say, is having, you know, multi-layered security available that covers your entire network, your data as it's in motion, your data in rest and at the various points throughout your network.

Dr. Dan Gerstein:

Great, yes, please.

Audience:

Anthony Serrano, Mantech. Question for Mr Basham. You mentioned things like data throttling, data protection, dynamic data, things that AI could potentially solve. And I'm asking you because you're in operations, so we know at the operational level Knox and Sox, collaboration with other departments, DOD, academia is well-intentioned and planned, but they're at the bottom of the list because you've got operations and tactical stuff going on. Do you see a renewed prioritization of that collaboration with DOD and other industries to implement AI for computing at the edge?

Craig Basham:

Thank you for the question. We are currently connected well. So when we started with the NOSC at the DHS level one, we started really working internally with the DHS and pulling all that component NOC and SOC the traditional ones back into one location and really porting that out, working with CISA also on their threat hunting and everything really started bringing it out across the major departments of the federal government, whether that's DOD, whether that's DOJ and the other areas. So we do work quite often with the other ones to collaborate to find out what they're doing but also getting that threat information across.

Dr. Dan Gerstein:

We have another question here.

Audience:

Huani from IBM Consulting. The question for David On the TSA side sounds like TSA already started doing some pilot with thousands of iPhones that you mentioned. Would you be able to share some of the primary use cases that you are looking at? Are you just focusing on the efficiency or maybe some transformative capability, with the edge devices being available? Thanks for the question.

David Earle:

Yeah, so we gave out those devices to our supervisory transportation security officers In collaboration with CBP. We allow them to use the CBP One app so when people are coming across the border they can be identified and that's on their phones and they can use that at the checkpoint. We started at the border, but now it's been scaled to the entire country so all the supervisors across the country have iPhones to be able to use that app and then, to your point, be more efficient. You know they can get data that they can relay down to their subordinates at the checkpoint.

Dr. Dan Gerstein:

Any other questions? We have time for one more.

Audience:

Great.

Audience:

Thank you, hi Jeff Smith, with Vintera. My question, outside of the devices themselves and the infrastructure to connect them, is the business systems to manage sort of extended operations around more devices at the edge, the asset management, the logistics management to replace them, the spares, basically the business systems to enable that compute at the edge to be maintained and resilient. And then maybe a second part of that question is how is it different for something like TSA, which might have fixed facilities, versus FEMA or even CBP that might have to have dynamic situations?

Dr. Dan Gerstein:

So who wants to start? That is an easy question. Cio, you want to?

Charlie Armstrong:

Yeah, I'll just say that we have a fairly robust logistics process for no-transcript. We try to keep it no less than 20% availability, so we've got a very good process in place for doing that today. The challenge, though, does become around some of these remote areas and I talked about Guam, I talked about Maui Alaska certainly are areas where it's difficult to get stuff in and out of in a rapid fashion. So those will remain to be challenges and, if I can speak for my former colleagues at CBP, they also have a very robust depot kind of situation where they can get equipment set up. Sometimes the vendor is responsible for actually doing that work and making sure that there's uptime and availability.

Dr. Dan Gerstein:

Greg David. Anything to add?

David Earle:

I mean we at TSA mirror pretty much what Charlie said. There we have a pretty good logistics group that gets everything out the door for us. We get stuff back into our warehouse in Springfield and get it right back out. So you wouldn't think getting an IT would be into logistics, but it is a lot of that getting stuff back and forth.

Dr. Dan Gerstein:

Jim, how about you Anything to add on that question?

Jim Westorp:

Anything about resilience on the edge as the last question, so, just in terms of the question about back office, I would say we spend a fair amount of time and effort for ourselves being able to interface with software and infrastructure planning tools that are northbound or above us. So there are a lot of standardized interfaces and specific application APIs programming interfaces that are designed to let the users or the management systems figure out what network assets are available if there's any problems with anyone, and make that visible so that, from a logistics perspective, they know what network paths are available or even what devices are in use right now or need service, et cetera.

Dr. Dan Gerstein:

Well, it's clear we could go on a lot longer, but Megan wouldn't invite us back, so we're going to adhere to the schedule. But how about a round of applause for the panel? Terrific, perfect.