Improving the Current Fleet

Guest: Robert Klueg, Division Manager, Developmental Test and Evaluation, Transportation Security Laboratory

Host: John Verrico, Chief of Media & Community Relations

 [00:00] Dave: This is Technologically Speaking, the official podcast for the Department of Homeland Security, Science and Technology Directorate, or S&T, as we call it. Join us as we meet the science and technology experts on the front lines, keeping America safe.

[00:00:00] John: Hello and welcome to this episode of Technologically Speaking, the official podcast of the Science and Technology Directorate for the Department of Homeland Security, where we get into the weeds on science and learn about the personalities of the brilliant minds that are working here at the Department. I'm your host, John Verrico, and I am delighted to introduce my guest today, Robert Klueg with the Transportation Security Laboratory, who's doing some absolutely amazing work. And Robert, I know we don't get this chance to see each other very often. It's been quite a while since I've had an opportunity to visit the lab, and to see you, so I'm so glad that we get to connect, today.

[00:00:54] Robert: I appreciate it, John. thanks for having me. And I really look forward to this opportunity to let you know some of the good work, that we're doing, here at the Transportation Security Laboratory.

[00:01:01] John: Yeah. You know, it's an amazing place, out near Atlantic City and it is, really a phenomenal facility and all of the amazing things that you do there. I'm not going to get into the weeds on it, I'm going to let you do that. But, first Rob, let's talk a little bit about you. how long have you been with the Department now?

[00:01:21] Robert: So, I've been with the Department for about 12 years. but I was at the Transportation Security Laboratory here for over 23. So, I spent half of that time as a contractor, working in some of our characterization labs that we have here. And then in, 2010, I became a, a federal employee with the Department.

[00:01:41] John: Outstanding. Congratulations for that. So let's talk a little bit about the lab, because you talk about the lab being around as long as it has, and we know that it was not part of the Department of Homeland Security originally. It actually was established prior to that. Can you talk a little bit about that?

[00:01:58] Robert: Sure. we were actually part of the Federal Aviation Administration, and when we were originally founded, it was, as a result of the Lockerbie bombings, and there was a thrust to improve the security posture in the United States. And the FAA was the lead organization in that sense for aviation. So, we were founded as the Aviation Security Laboratory, and we, supported all the development of the baggage screening systems, the checked baggage screening, all the metal detectors and x-ray imaging systems that you have seen over the years, have all come through this laboratory. And then after 9-11, we became part of the Transportation Security Administration for a few years until the full Department of Homeland Security was stood up a few years later. And then we were, brought into the newly created Science and Technology Directorate where we've remained ever since.

[00:02:53] John: That's outstanding and you know, you point out Lockerbie and I was really excited that we're recording this episode at this time because as you know, the suspected bomber from that incident was recently arrested and put into custody. It's amazing to think that was back in December of 1988 and they finally caught this guy. How did that make, I can imagine just a bit of celebration going on around TSL?

[00:03:21] Robert: Oh, we follow those developments very closely. It's ingrained, obviously, in our mission, right? To make sure that the traveling public is safe and can fly and move around the country and the world conveniently. And to see these bad actors, just taken off the world stage is just, another reminder of, how important our job is from day to day and that we're all kind of doing our things behind the scenes to keep everybody safe and, keep everybody moving.

[00:03:46] John: Could you talk a little bit about that Lockerbie bombing? Just a little bit just to, to give folks who may not be familiar with it, especially since it was so long ago, exactly what had transpired? 270 people perished during that incident.

[00:04:00] Robert: Yeah. There was a military type plastic explosive, hidden in an electronic device. I believe it was a basically like a boombox, you know, a radio at the time. And, was, put on board an airplane and detonated, over Lockerbie. As you said, you know, everyone perished in that event. And again, that, that type of scenario, was the impetus for. expanded screening, of luggage and baggage and cargo and we've developed technologies initially to kind of counter that type of threat. But as the adversaries become more sophisticated, they've learned kind of how to adapt.

[00:04:36] We've had to adapt so even though you may think that, oh, we’ve kind of, we have all the security out there. Hey, problem solved. given, just as you said, it was so long ago that, that particular incident took place. but the threat has just evolved. It’s continually evolves. And so, we make it our, mission every day to make sure that we stay one, step ahead of those adversaries.

[00:04:57] John: And we know things have, evolved so much and the incidents happen all the time. There was the shoe bomber. There was the underwear bomber, all of these kinds of things that keep you guys on your toes, trying to figure out, you know, what is the next threat? What is the next way they're going to try to cause damage on an aircraft?

[00:05:18] Robert: Yeah. and we, have really developed, over especially the last 10 years, a really good relationship with the intelligence community. We have, partnerships with the F B I and, the National Counter-Terrorism Center and, other groups, you know, worldwide to try and get the most recent intel and most relevant intel to make sure that we are keeping ahead of those potential threats. We have had several incidents over the years where we've gotten that intel. We've worked with our inter-agency partners very closely, mocked up devices, mocked up scenarios and gone through and tested those, IEDs, those improvised explosives devices or scenarios on our security equipment and made sure, hey, that's going to be detected or here's the data that we need to make sure that never gets on an airplane. You know, that's really been a great development. Like I said, over the past 10 years, we've really established a lot of good partnerships, with our intelligence colleagues.

[00:06:12] John: Before we get into the weeds of all of that, which I am dying to do because that's the fun of all this. Just a little bit more about you and how you, kind of got involved in this, this realm.

[00:06:26] Robert: Sure. My background's in, physics and so, I had done in undergraduate and graduate school, work on materials characterization with X-rays and some nuclear techniques. and so, I got really lucky to be honest with you that, the. Aviation Security Laboratory at the time. I, graduated in 1999. they stood up or were standing up, a explosive characterization lab, a bulk explosive characterization lab, and my skillset just happened to be aligned with that mission. And so, I was able, as a contractor to work very closely with, my government colleagues to essentially establish a capability where we could characterize explosives in their X-ray properties in such a way that we could look to differentiate those from all the other stream of commerce that, you know, maybe, you know, carried in cargo or people's luggage and that sort of thing. So that way we could really hone in on, on the materials that we, were really interested in that needed to be stopped and then not have to worry about, you know, high false alarms, nuisance alarms, things like that, that would inconvenience, the traveling public. So, my background technically aligned both with, you know, the interests of the government, but also that timing piece of having that lab stood up kind of at that time when I, was looking for a, a challenge in life. So, it actually worked out really well.

[00:07:44] John: How rewarding is it for you to know that you’re applying your scientific background to such an important mission?

[00:07:51] Robert: That's what it's all about. that is, the thing that we talk about here at the lab, often. It's how present we are with respect to the mission. We see our families, our friends, you know, travel all the time. You always see people, you know, businesses, reliant on being able to move around the country so it's a very real, mission for us. And, to be able to apply your technical knowledge to something that you're going to see it, you know, maybe in a year or two you're going to see that, that piece of equipment, that technology that you helped develop, and prove that it works to protect people. You'll see that in the airport, you'll see that at a train station. You'll see that guarding a, public space. You know, having that connection, that real connection to the technology to develop and to the people you're supporting, really motivates everybody here to do what needs to be done to get the job done.

[00:08:48] John: You just exude that enthusiasm and that’s why I love these conversations. So, tell me now about some of the work that you're doing now, and especially with, the Screening at Speed Program and how we're trying to really enhance the effectiveness and efficiency of our ability to detect, explosives and materials like that?

[00:09:11] Robert: You, your question hits the nail on the head because really what we're always moving towards is to try to be, efficient and effective, as you said. And the way we do that is, we try to increase the number, the types, the diversity of threats, as that adversary changes, in our detection portfolio. While trying to reduce the inconvenience, the nuisance, the false alarms that come along with scanning the general public. So, to do that in screening at speed. it's a challenge. I mean, you’re trying to actually increase, the, the number of things you're detecting. But to reduce that nuisance rate, is not trivial. And so, one of the things that we do to support, the screening at speed is to try and service the existing fleet of screening systems, especially personnel screening systems. So those are the ones that screen your body. there's a, an antenna that goes around and images your body and sees if there's any anomalies, anything you're carrying, concealed, things like that. They're already out in the airports and we're trying to continually upgrade those algorithms. Okay. So, one of the things that we do is, as we get that intel, or if there's an, if there's a, an incident somewhere, we work with our Screening at Speed, colleagues, John Fortune, Dr. John Fortune, who's the program manager and our TSA colleagues.

[00:10:33] John: I had the opportunity to interview him for one of these episodes as well.

[00:10:37] Robert: Yeah, and he's a great partner. We work very closely with John and his team. And we've been doing so successfully for years in, transitioning, products to the field. So, it's a great partnership. and we get that input from TSA, and we will come back here, create those scenarios, create those threats, and we actually have, for those personnel screening systems, we actually have a pool of mock passengers. These are normal everyday people that we bring in., and we can, work with them to put those threat scenarios, on them. Okay. and we have a institutional review board that goes through, because there, there's human in the loop testing and that sort of thing. So, there's lots of safety protocols to follow. But, we will go and actually run those folks through the system to collect data that would then train an algorithm to detect those threats.

[00:11:25] John: So, you'll conceal threat items on a mock passenger. Run them through the system and through your equipment to see whether or not your equipment can detect the stuff?

[00:11:35] Robert: Exactly. And if that threat is, so new, then we take that data and we provide it to the maker of that equipment so they can then update their library and their detection algorithms. So now that's now included in, their detection portfolio. So, we try to, utilize that real world situation. We don't want to be satisfied with just a cursory laboratory experiment, so to speak. We use real passengers. we look at their clothing at different times. We have different scenarios for different weather conditions. We do all this to make sure that we really try to encompass what the real world would present to one of these systems. And that way we have confidence that they will work. So that's what we do, to support a lot of the fielded systems. Screening at Speed as I'm sure Dr. Fortune had spoken with you about is really forward looking and trying to reduce the divestiture that is required for the traveling public. You know, you have to go and take off your jacket, empty your pockets, take off your shoes. We want to move to a concept of operations that doesn't force the passenger to do all that. Leave things on, leave your jacket on, leave your shoe on, that sort of thing.

[00:12:50] John: By reducing those inconveniences, it actually makes the whole process more palatable for people. Yeah.

[00:12:57] Robert: Yeah, absolutely. Yeah, and the outcomes for the public and for the security posture are very positive because if you can have people leave on their jackets, their shoes and what not, and actually get that flow moving faster, it affords more time for if there is really a significant issue that comes up, you have a potentially more time there to do a more adequate inspection. And maybe bring other tools to bear to give you a better confidence in whatever your result is. there's a lot of room to be able to improve the entire experience. And so, we're working with them to, look at, for instance, prototype shoe screeners where you can leave your shoes on. It could potentially be integrated into the floor of a, body scanning system in what we call an Advanced Imaging Technology System, or A.I.T, to be able to just have the person be scanned, you know, the body and the shoe at the same time, so that way there's no divestiture of something like that. We're also working with, the screening at Speed program to move to higher resolution body screening systems. Okay. And so, both of these systems are developed in partnership with, Pacific Northwest National Lab, P N N L. and they're developing a millimeter wave. They're great and they've, over the years, they've developed a lot of technology that's been transitioned into the field. The current crop of A I T systems I had mentioned before, is at least in part based on technology that was developed, at P N N L and now they're developing the next generation, we're working with them as well.

[00:14:32] John: Wow, that's just absolutely amazing. And now we're advancing, toward, kind of a self-screening kind of a system.

[00:14:40] Robert: Yeah, and that's really one of the ultimate goals is to be able to have the passengers, in essence, process themselves. Some of the things that Screening at Speed is looking at is both kind of a, more of a walkthrough or, you know, less intrusive, screening modality. And it's almost like, if you remember the Arnold Schwarzenegger movie, Total Recall, people like to talk about that wall they have where, you know, they're, you just walk through and then it, and then I think he was carrying a, gun or something, and then it just alerts on that spot and there's that, there's, you know, no, nobody interacting with that person to get that alarm. And so, it's that kind of, idea at least. And so, there are several screening technologies that we're, we have here at the lab. and one of them is a real-time imaging system from, P N NL that is very reminiscent of, that Arnold Schwarzenegger, movie screen. It basically is a real time. millimeter wave imager. You, walk through it and it provides, a very similar image to that movie image. And it doesn't have an algorithm. So that's the thing where we come in, where the TSL will work with them, we'll provide them threats. We'll provide weapons, knives, guns, things like that. Anything that's prohibited. And we’ll, work with them and our mock passengers to collect all that data that we can then provide not only P N N L, but third party algorithm developers, maybe other national labs that would have insight into how to create a robust and efficient algorithm. And, we'll provide all that data, to them, to those parties, and they can develop the best in class detection algorithm for that type of technology. That's at the early stages, but that's kind of one of the things that, we, think will be the future, to be honest.

[00:16:22] John: You know, we're talking about, all of these images and the capabilities of these various types of sensors, and imaging systems that can see all of this stuff. But the actual Transportation Security Officers don't actually see these images, right? They go into a computer and the computers use these algorithms to define what it is that they're seeing, and then they can then alert back out to the officer?

[00:16:46] Robert: That's correct. We use the term avatar. the Transportation Security Officer doesn't see, as you said, any of the level of data that I'm referring to. So, this is all done computationally and then there is a generic, you might want to call it a stick figure if you want to have that picture in your head. But it's just a very generic cartoon of a person that just will have the prototypical red, flashing red light where there's, a potential problem and alarm. And so, none of that, that imaging is saved. It's not, it's not viewed by anybody. And so that way you can, kind of, protect people's privacy and, any concerns they have there. But give enough information to the Transportation Security Officer where they can resolve an alarm. On top of that, we're also working with the screening speed folks to develop alarm resolution technology. One of the things that we want to minimize is that pat down, right? Is that interaction with the public. No one likes to, to be pat down, no one wants to be stopped, that sort of thing. So, we want to minimize that. So, if something does alarm. We're working with manufacturers and equipment developers on, for instance, hand wands where you could have almost like if you've, ever been wanted with a metal detector, a hand metal detector.

[00:18:05] John: That's great. So, what's coming down the pike? What can people expect to see, happening as this kinds of, these kinds of technologies continue to evolve?

[00:18:15] Robert: The end goal is that you, could actually see less in the sense that you have an architecture. and if we get to self-screening where you may be able to just scan a boarding pass and then, and put your bag on a conveyor and walk down a corridor and, and that screening is taking place as you're moving through that system. And then you meet your bag on the other end. Okay. And it's been, screened and passed through the system, and you grab it and move on your way. That's the ideal end goal. Along the way, you'll see iterative improvements. We're already putting in. TSA is already putting in, cat scanners, CT systems at checkpoints. that as those systems progress, in their training, algorithm development, training, that will become a very proficient process, for just putting your bag on a belt, having it screened and running it through a, an automated screening lane.

[00:19:11] John: Could you go into the weeds a little bit on, on how these CAT scans work?

[00:19:15] Robert: Sure. Absolutely. So, it’s just like the medical scanner that you would, if you had a CT scanner, a CAT scan, it's an X-ray image. They acquire approximately a thousand images as on a, as a detector and a source, rotate around your bag and they, the computer will then reconstruct a cross section of your bag. And then we have algorithms that will look at the material properties of those objects in the bag. And then either there'll be an engineered algorithm that'll say, well, if it has these properties, they're associated with an explosive, we need to alarm that. And then that can get moved to some sort of a secondary screening or an alarm resolution. Or we're moving towards machine learning based detection algorithms for these CT systems. And that the machine learning algorithm may say, yes, this matches the profile of explosives that I've been trained on. Or it may say, oh, that's a gun, that's a knife. You know, in your carry on, that's prohibited, and it will alarm. And so, we'd like to get to a point where those CT systems only have a person in the loop as an alarm happens, so you don't necessarily have a person viewing each individual bag like you would with kind of that standard projection X-ray that's there now, that it would only alert somebody if it thought something was in the bag.

[00:20:41] John: So, Robert, when you're running through all of your various testing and the machine learning and you're training basically these, sensors and these devices to be able to, not only detect, but then to define what they're seeing, you're using all kinds of data. And you, we've talked about how you're doing those real-time, hands-on, types of things and, giving a, an anomaly object to, a mock passenger and saying, here, conceal this somewhere on your person and let's see if we can find it. And then you're putting things in baggage and randomly trying to see if you can find it and all that great stuff. But there's all the kinds of data that you're using as well. And I've heard the expression before called synthetic data. Let's talk a little bit about that and what that means.

[00:21:26] Robert: Sure. You know, I had mentioned a few times already about, machine learning and artificial intelligence, being the next wave in detection algorithm. systems and that's going to be the big, you know, the big advance. and one of the concerns that we have as the government is, those algorithms need to be trained and they need to be trained with a lot of data, factor of a hundred or a thousand or 10,000 more than we provide now. As that group that does a lot of that, that vendor algorithm training, we need to figure out a way to provide the width and the breadth of data that would be required to, to train, machine learning algorithms. And one of the ways that we're looking at doing that is with synthetic data. Okay? Data that is not necessarily collected physically on a system but is generated through a computer that mimics the properties that system would have measured. we can generate images computationally very quickly, fairly inexpensively and provide those to the vendor community, to be able to make quick algorithm updates and things of that nature. The other, benefit to synthetic data as well is, currently, as I had mentioned before, our adversaries are coming up with, new explosives. New mixtures. They take whatever they can get. We call them homemade explosives or improvised explosives, and they can, you know, they can take on many forms and, for us as you know, scientists, engineers, we’re not necessarily as ignorant to the risks as our adversaries may be. And so we want to do everything safely, effectively, and repeatably. And that can be, that's a challenge to do that with some of these homemade explosives. And one of the ways that we could potentially improve that process is by not having to make them, but, physically, but we can make them virtually, right? So it, so anything that would reduce burden of having to kind of create these improvised explosives, would help us in our jobs.

[00:23:24] John: So it’s not like a kid with a chemistry set who you know is playing around and mixing things together to see what happens and blows up the garage. You're kind of doing this all, virtually, instead to see what would happen by combining elements.

[00:23:38] Robert: That's the goal.

[00:23:39] John: That's amazing. I don't even understand how you think this way, how you come up with ways to think about the fact we rely on computers to do so much thinking and processing for us. But you guys are so ahead of this because you're thinking of how to train the computers to train themselves. To be able to do all that thinking and processing. It's just, absolutely phenomenal.

[00:24:03] Robert: Yeah. And that's a very good way of putting it and, you know, we're borrowing a lot of the ideas and things that have gone on with the Googles of the world and kind of those cutting edge technology companies. But, you know, our mission is so unique that we can't just take off the shelf processes. Generally, we're a small community. The explosive detection community, the threat detection community is a fairly small community. So, we try to kind of, you know, take the nuggets where we can from, from some of those bigger industries. But we always have to make them our own at some point because of the uniqueness of, and the challenge that comes with what we do.

[00:24:37] John: As usual, I always learned something every time I talked with you and with, your fellows up there at the Transportation Security Lab and throughout the Science and Technology Directorate, which is why I love doing this podcast. It gives me really an opportunity to learn so much, every single day. I realize, you know, how feeble a brain I have is compared to, to your guys with these super minds, it's these superpower minds. I'm just going to ask you an off the wall question because we're talking superpower. I just, when I say superpower, you know, if you had another superpower other than this super brain of yours, what would you want that superpower to be and why?

[00:25:16] Robert: Oh my, that's a good question. I guess if I had to, if I had to pick one, I don't know. I guess, I guess being, able to fly, what about that one? I'm not a patient person when I'm in a car. I'm not a patient person. you know, to get around, I guess. So, I guess if I had to pick one that's what it would be. I’d, like to be able to go from point A to point B as quickly as possible.

[00:25:39] John: So here you are actually helping the world be able to do that, be able to fly as quickly as possible. And I didn't plan that remark, but I guess you, you know, that's a good point. I guess we're all looking to do the same thing one way or the other, yeah. It all kind of comes back to it, right? You are perfectly suited for where you are and what you're doing. Thank you so much. Robert Kleug with the Transportation Security Lab, here at the Department of Homeland Security Science and Technology Directorate. And you have been listening to Technologically Speaking the official podcast of S&T, where we get into the weeds and behind the scenes and into the brains, and into the hearts of the people who work here. And so, it's always a delight, to do this program and to chat with, the folks around the Directorate. Robert, thank you for taking the time to do this.

[00:26:26] Robert: Oh, John, it was a lot of fun.

[00:26:28] John: This has been Technologically Speaking, the official podcast of the DHS Science and Technology Directorate. To learn more about S&T and find additional information about what you heard in this episode, visit us online at Scitech.dhs.gov and follow us on social media at DHS SCI TECH. Thanks for listening.