You are here

Episode 4: Inspection & Direction

Dan Gillins Postdoc Researcher Eric Becker
Dan Gillins Eric Becker

In this episode, we find out how researchers are utilizing unmanned autonomous vehicles (UAVs) to gather data.

Dr. Dan Gillins, affiliated faculty member, along with his brother Matt Gillins, a graduate student at Oregon State, is conducting research on the feasibility of using UAVs to conduct bridge inspections. Gillins’ research seeks to determine if UAVs can help defray some of the cost and eliminate some of the risk involved with traditional bridge inspections. 

COE postdoctoral researcher Dr. Eric Becker developed a UAV-mounted Radiation Compass that maps radiation sources autonomously in real time. The Radiation Compass weighs only 650 grams and can be built for $3,000 giving it a major advantage over current detectors.   

Bonus Clip

Listen to Matt Gillins' full thesis defense: Unmanned Aircraft Systems for Bridge Inspections: Testing and developing end-to-end operational workflow.



DAN GILLINS: There were days where we used to joke, I used to joke with my field crew: "Wouldn't it be great if had a, a jetpack

[jet pack blast]

that we could strap on our backs and fly to the next point that we needed to survey and collect the data that we needed?" And, don't get me wrong, I really enjoyed the field work, but there were, there's always those bad days where I wish that technology was available. UAVs, of course, is not exactly that technology, but it's pretty darn close. In fact, maybe it's even better.

[MUSIC: Eyes Closed Audio, The Ether Bunnyused with permissions of a Creative Commons license.]

KRISTA KLINKHAMMER: From the College of Engineering at Oregon State University, this is Engineering Out Loud. Now, maybe someday, we'll be talking about using jetpacks to collect data, but in today's episode, we're talking about research efforts at Oregon State that use Unmanned Aerial Vehicles, or UAVs

[car locking up brakes and screeching to a halt]

Hold up-- before we go any further, let's talk terminology. There are many different names used to describe an aircraft with no pilot on board. People often use UAV, UAS, drone, and multicopter interchangeably. In this episode, you'll hear it referred to as a UAV or UAS, so I just want to distinguish between the two. An unmanned aerial vehicle, or a UAV, generally refers to the aircraft itself. An unmanned aircraft system, or UAS, refers to the aircraft, but also includes the ground control and communications units associated with it, including the human operator. With that clarified, let's go on. The clip you heard in the opening was Dan Gillins. He's an affiliated faculty member at Oregon State, and currently a geodesist in the Observation and Analysis Division of the National Oceanic and Atmospheric Administration. In the first part of this episode, you'll be hearing about research he conducted using UAVs to inspect bridges, while he was an Assistant Professor of Geomatics here at Oregon State. You'll also learn that this work tends to run in the blood in some families. 

MATT GILLINS: I just want to echo his sentiments of thanks for all of you to be here today, to hear a bit about my research in unmanned aircraft systems, and how we can apply them to augment bridge inspections. 

KLINKHAMMER: That's Dan's brother, Matt Gillins, introducing his master's thesis defense, also here at Oregon State University. And here's a little more context from Matt's thesis defense on how deep this family tie runs. 

M. GILLINS: Before we get too far into my presentation, I kind of want to take a step back and think about bridges. And it's kind of where I started in the interest of civil engineering, from a young child. The bridge shown here on the screen is a beautiful bridge located in central Utah, central-southern Utah. It's called Hell's Backbone Bridge, and it was actually designed and engineered by my father, who was a structural engineer his entire career. 

KLINKHAMMER: So, brothers Dan and Matt, inspired by their structural engineer father, have spent a lot of time researching the use of UAVs

[UAV flying]

for bridge inspections. Some of the research Dan is conducting has been funded by PacTrans, the Pacific Northwest Transportation Consortium, and also the Oregon Department of Transportation. The goal is to fly different bridges to determine what the technology is capable of, and how it could benefit some of these transportation agencies. 

M. GILLINS: So, for bridges, the Federal Highway Administration actually mandates that every bridge in a way highway system road needs to be inspected at least once every two years. And so that's actually a pretty, that actually requires a lot of work. In fact, the Oregon Department of Transportation spends, on average, somewhere around four million dollars a year inspecting bridges in Oregon, and helping run that program. So, that usually involves a lot of things-- it may involve hiring specialists who are climbers and engineers to go with harnesses, and hook onto a bridge, and climb it, in order to see anything, any kind of defects that may be on the bridge. It may involve mobilizing these large trucks, or what we call them "snooper cranes," which they would have to close one lane of the bridge, put the crane on the bridge, and then the inspector will stand in a bucket, and then kind of be dangled out underneath the bridge to do their work. So, with the UAV, the idea is pretty simple, actually; we thought this could be a technology that could be used where we don't have to put the inspector in that harm's way.

 KLINKHAMMER: Matt also emphasized the safety benefits in his thesis defense. 

M. GILLINS: So, right out of the bridge inspector's handbook that the Federal Highway Administration created, it says in it, "bridge inspections are inherently dangerous."

KLINKHAMMER: And here's Dan again. 

D. GILLINS: Really, our focus is evaluating what's the feasibility of the technology, what can we do, and also what can't we do? Because there are certainly things that just can't be done with it, too. It's not a miracle technology that'll solve all the problems. 

KLINKHAMMER: Dan went on to explain that most bridge inspections are done visually, and many inspections require the inspector to be at arm's length of the bridge. When flying a UAV of course, the inspector is not at arm's length, so the goal of the research is to see if the visual imagery collected from the UAV can simulate an arm's-length inspection. During a trial run at a bridge in Salem, Oregon, they flew a low-end commercial drone along a bridge about five to six feet away from the bridge, capturing high-definition video. The video was broadcasted real time to an on-site inspector who reviewed the footage and noted if there was anything else he wanted to take a closer look at. 

D. GILLINS: Another thing you can do when you're flying a drone is actually program it to fly a certain path. So, you specify, you know, 'fly to these different points, and capture imagery,' and it will trigger its camera automatically, so it's really nice and helpful for the pilots, so they can actually put in the flight plan and then hit 'go,' and it takes off, flies the plan, taking the pictures in a very systematic way. Well, why would this matter? Well, it's nice, because you can actually, if you do this correctly, if you have enough overlap between each picture, we have software where we can actually take these images and build 3-D models. 3-D maps, 3-D models of either a site, or a bridge, or a road, or any of these types of things that engineers would be interested in. So, I think, in terms of data and engineering, this could be really helpful data--it gives them a 3-D way to look and visualize everything in three dimensions if they're designing, if they're doing some engineering designs, or if they're trying to, you know, simply document what the existing features are in space--they have this, and we can provide that with the, with the types of flying and data collection we can do with the UAV. 

 KLINKHAMMER: Dan continued explaining additional benefits of using UAVs for bridge inspections. 

D. GILLINS: So, for the bridges, um really it's a great way to document and have a digital record of the conditions of the bridge. So, in addition to being able to, in real-time, view elements of the bridge and try to see, you know, whatever issues, whatever defects may be on it, and, say, for example a bridge inspector, you know, wants to see a loose, look for loose bolts, or rust, we can kind of see and capture that in the imagery with the UAV, or that's what we're trying to do. But, in addition to that, the images that we capture can be archived and stored, and so if, for example, they then go back to the same bridge two years from now, and do a similar inspection, they now have what was there, versus what's there today. And they can see if there's any changes or how rapidly things are changing, and how rapidly is a crack on the bridge growing, or how rapidly is the rust forming on the bridge, and they can get that kind of information. And so, that I think is a really, really huge benefit--it's not just based on the memory of the inspector, but there's actually another digital record of the site they could use for detecting change.

KLINKHAMMER: While Dan is pleased with how the research is going, he also talked about the challenges.

D. GILLINS: We've done this, it's gone well. There's definitely some challenges; one of the challenges we have are, one of the big challenges that we have, are that many of the drones or UAVs we fly rely on GPS to help the pilot--and it's really useful, that way if the pilot gets confused, or if there's a gust of wind, or any issue that goes on, they can simply let go of the joysticks and the drone will just hover in place. And that's really useful, but the problem is is when we get close to a bridge, the GPS signal starts to get degraded, because the bridge is blocking the signal. And it's a real problem, actually, if we want to go under the bridge, which is typically what an inspector wants to see. And so that's something we're working on, is trying to use other sensors instead of just GPS to help with these kinds of inspections.

KLINKHAMMER: The possibilities this technology will allow are exciting, and the researchers are still testing the limits of what can be done using UAVs.

D. GILLINS: We want to see, when we're testing, using current technology, what are the capabilities, and what are the limitations? And simply trying to inform agents, transportation agencies, if they should invest in the technology, or not, so that they're more informed, so that they can make better decisions. And, if, in fact, they are pleased with the capabilities, then maybe they could then, we would then, provide them some information and assistance with how they could implement that in their inspection program, in a safe way. So, we're really interested in, you know, what can we do, and what can't we do? There's certainly things we can do-- we can, we can into positions that are otherwise hard to climb to, especially anything like a tower above the bridge, or even, in some cases, beneath it, so that's really, really helpful. We can see things that we want to be able to see, in many cases. There are some limitations that we've noticed right off the bat. Like, one thing we definitely can't do is do any scraping or probing on the bridge, which is usually something an inspector would want to do, is be able to wipe the bridge or maybe knock on it, or get a feel for what's, you know, if there's some rust or some other material on the bridge, be able to wipe it, and get it and look at what they want to see. We can't do that with the UAV; it would just be primarily visual.

KLINKHAMMER: Dan doesn't intend UAVs to ever completely replace a human in conducting bridge inspections, but he did talk a bit about where he thinks the technology is headed. 

D. GILLINS: Well, first off, I guess I should say I think we're going to see UAVs become more and more popular. We're going to see engineering offices that have them, that are using them to do routine projects, and so I think that's something we're going to have to get used to. But also, I think I can see we're, the direction we're heading is now putting additional sensors on the UAVs, so whether it's, instead of just a camera-- some of this is already happening right now, where you can put an infrared camera on it, or a light-r sensor on it, but I think you're going to see other sensors that are put on them. Another thing that I think that is really important, that will help and assist the pilot, so that the things can be done more safely, is, I think sensors are going to be added to help sense and avoid obstacles. Some of this technology's already available; actually, on one of the UAVs we fly, it has ultrasonic sensors on it, so we when we get close to a feature, it will start beeping at us, and warning us that we're about to hit it. We can even use it to actually lock on to an object and hold an off-set distance, and help us. So, it's all very helpful for avoiding obstacles. But I think, as the technology progresses, it's going to get to a point where it's very smart, and can actually, on its own, sense an obstacle and avoid it, and fly around it for you. And I think when that happens, then it's going to become even more prevalent. 

[MUSIC: Eyes Closed Audio, The Ether Bunnyused with permissions of a Creative Commons license.]

KLINKHAMMER: Dan and Matt's work on bridge inspections is just one example of how Oregon State researchers are using UAVs. Stick around for the next part of this episode, where you'll learn about a UAV-mounted radiation detector. 

[MUSIC: Eyes Closed Audio, The Ether Bunnyused with permissions of a Creative Commons license.]

PRESIDENT BARACK OBAMA: Fortunately, as I said this morning, no terrorist group has yet succeeded in getting their hands on a nuclear device.  Our work here will help ensure that we're doing everything possible to prevent that.

JENS ODEGAARD: That’s President Barack Obama speaking at the closing session of the 2016 Nuclear Security Summit. Tracking and securing nuclear substances and radiation devices is key in the ongoing effort to keep people like you and me safe and ensure that the use of the atom is restricted to peaceful purposes.

I’m Jens Odegaard and today on Engineering out Loud I talk with one of those involved in this work. Dr. Eric Becker holds a Ph.D. in Radiation Health Physics from the College of Engineering’s School of Nuclear Science and Engineering, where he’s currently a postdoc researcher. His recent work includes the invention of a UAV-mounted radiation detection device—a device he’s dubbed the Radiation Compass.

ERIC BECKER: So the radiation detection work that I do is primarily focused on work supporting non-proliferation and treaty verification and national security. SO, if someone say is trying to smuggle materials or even a small nuclear device in the country my work is aimed at stopping that from happening.

[MUSIC: Himalayaused with permissions of a Creative Commons license.]

The Radiation Compass is actually a set of 16 detectors that can tell in which direction a source of radiation is located. And, we intend to use this detector by mounting it to the underside of a small unmanned aerial system. And we can use this for tracking down perhaps a nuclear device that has come into the area illegally or you can use it to sweep

[broom sweeping]

a street to make sure nothing is being hidden in a store or something.

ODEGAARD: The entire Radiation Compass in its current iteration only weighs 650 grams or about 1.5 pounds and it fits on the underside of a hobbyist size drone—something you’d see filming a wedding for instance. The 16 detectors are arranged in a circular pattern about as big around as a dinner plate. This 360 degree pattern allows it to pinpoint in real time which direction a radiation source is located.

BECKER: each detection element of the Radiation Compass uses a bismuth germanate scintillator, which is a material that produces light when radiation interacts with it. Radiation comes in, it produces light in the scintillator material, and then the light is captured by our photo sensor and converted into an electronic signal. That electronic signal flows to a readout module that is reading out all 16 of our detectors in parallel. When we get an interaction like that that we can detector that produces light for our photodetector to see, we call that a count.

ODEGAARD: The total number of these interaction events, or counts, over a period of time is plugged into a statistical estimation method called matched filtering that’s common in electrical engineering.

BECKER: What this method does is, is it compares the measured detector response in aggregate from our complete detection system of 16 detector elements and it compares that to a library of results that we obtained by simulating our detector using a radiation interaction simulation package called MCNP.

ODEGAARD: Using this Monte Carlo N-Particle Transport Code, Eric created a library of 80 responses--one for each 4.5 degree interval of a 360-degree circle.

BECKER: Based on which response receives the highest value as a result of this process, that's the direction that this method is telling us is the most probable direction the radiation source is located in. 

ODEGAARD: All this information is relayed wirelessly back to an operator who can be well clear of harm’s way. This, along with the small size and directional capability give it a big advantage over other detection devices currently in use. Eric explains:

BECKER: So a lot of law enforcement and homeland security personnel use large volume detectors. And all they're trying to do is determine whether or not a radiation source is present. For example, they will

[large diesel vehicle driving]

drive a big vehicle-mounted detector back and forth up and down the street and try to determine whether or not there is a nuclear device behind the storefront or something like that. These detectors are simpler than the Radiation Compass. They can't tell what direction the radiation is coming from.

[large diesel vehicle driving]

These vehicles actually have to drive up and down the street twice, once up, once down because they need to be able to tell which side of the street the radiation source is coming from. With the Radiation Compass you just have to fly a drone down the street once because it's able to tell you direction it will be able to tell you which side of the street the source is on. So that's one example of an advantage.  

ODEGAARD: In addition to vehicle-mounted detectors there are other drone-mounted detectors, but they too have their downsides.

BECKER: Fortunately at least currently these devices are a lot heavier than our detector, so it would require a bigger drone to actually fly them around. And mostly because of the detector material that they use they are also much, much more expensive. We estimate that the Radiation Compass could be built for around $3,000 for parts alone. But these other detectors, there's actually an example of an imaging detector that's mounted on a larger drone. This particular detector that I'm thinking of would cost almost $150,000 for just the detection material alone to speak nothing of electronics or the drone itself.

ODEGAARD: Size, cost, and directional capability aren’t the only things the Radiation Compass has going for it. Eric, in fact, envisions a future in which the compass system could be used in conjunction with other devices to both map and identify radiation sources.

BECKER: So the Radiation Compass is actually, in its current state, used purely for finding what direction a source of radiation is. However in the past, our lab has developed a very small wireless spectrometer that could easily be integrated into this system. The idea is that when this Radiation Compass combined with a drone does 

[UAV flying]

find a source it can land directly on the source and then the small spectrometer can be mounted to the underside of that and take an energy measurement and identify the source.

ODEGAARD: Together this unmanned system could much more efficiently scan an area than humans could, while greatly reducing human risk factor.

BECKER: The Radiation Compass mounted to this unmanned drone

[UAV flying]

can actually fly ahead of law enforcement or soldiers or emergency responses personnel and take readings. We talk about the main feature being that it can find sources of radiation all on its own, but it can also be used to map an area so you can tell first responders where not to go, where it's too dangerous, where they might only need a hazmat suite. By doing that we would essentially be trying to save people from being put in harm's way. In the future, we'd love to see the Radiation Compass become a tool that can be used by law enforcement, by homeland security, by emergency responders to both find radiation sources that maybe are being brought in by malicious forces or find radiation sources that have simply been misplaced. We see it as a tool to help people find sources of radiation to help make everyone else safer.

[MUSIC: Himalayaused with permissions of a Creative Commons license.]

ODEGAARD: As a tool for finding and identifying radiation sources, the Radiation Compass then becomes part of the mission laid out by President Obama in a 2009 speech in Prague.

OBAMA: Some argue that the spread of these weapons cannot be stopped, cannot be checked -– that we are destined to live in a world where more nations and more people possess the ultimate tools of destruction. Such fatalism is a deadly adversary, for if we believe that the spread of nuclear weapons is inevitable, then in some way we are admitting to ourselves that the use of nuclear weapons is inevitable. Just as we stood for freedom in the 20th century, we must stand together for the right of people everywhere to live free from fear in the 21st century.

[MUSIC: Himalayaused with permissions of a Creative Commons license.]

ODEGAARD: This episode was produced by Krista Klinkhammer and myself. Mitch Lea was our audio editor. Our intro music is The Ether Bunny by Eyes Closed Audio. The song you’re hearing now and earlier in this Radiation Compass segment is Himalaya by The Fish Who Saved the Planet. Both can be found on Soundcloud and were used with permission of a Creative Commons 3.0 Licenses. If you’re interested, you can check out Matt Gillins' full thesis defense Unmanned Aircraft Systems for Bridge Inspections: Testing and developing end-to-end operational workflow in the episode show notes on our website engineeringoutloud.oregonstate.edu. Catch you on the flipside.