Detection dogs: Sniffing out explosives, invasive pythons and... COVID-19?

Tiny Matters

In this episode of Tiny Matters, Sam and Deboki go on a detection dog deep dive. These are dogs who can sniff out basically anything, from explosives to invasive pythons to diseases like cancer and COVID-19. How are they trained? Are certain breeds better at recognizing certain odors? And, come to think of it, how do any of us smell things?

Transcript of this Episode

Sam Jones: [Sound of dog sniffing] Right now you’re hearing the sounds of a ferocious canine making her way through the wilds of … my living room. It’s my dog Yuca, who has her face in the shag carpet, probably looking for crumbs from a treat that she was chewing earlier.  

[Dog sniffing] And that’s Tzatziki—or Ziki—my other dog. She’s also very enthusiastically sniffing something out.

Why am I talking about my dogs (other than the fact that I love talking about my dogs)? Because this episode is all about dogs who are trained to sniff out a lot more than treats.

Welcome to Tiny Matters, I’m Sam Jones.

Deboki Chakravarti: And I’m Deboki Chakravarti. Today on the show, Sam and I are doing a deep dive on detection dogs—dogs who can sniff out explosives, drugs, maybe even COVID-19.

But before we talk dogs, we need to talk about smell. How do we smell things? What exactly are ‘smells’? Sam, knowing your research background, I’m gonna let you take this one.

Sam: Yeah this is, or I guess used to be, my jam. My graduate work focused on the olfactory system, which is basically our smell center.

In our noses, and in dog noses too, there are cells called olfactory sensory neurons. And each of those olfactory sensory neurons has proteins sticking off of its surface called receptor proteins. That’s where odor molecules bind.

So when we, or a dog, take a sniff, we’re sampling hundreds or thousands of odor molecules. Odor molecules could be a lot of things: vanilla coming from the perfume on your wrist, some gross garbage-smelling compounds evaporating off of a pile of trash, or… chemical compounds coming off of an explosive. And when those molecules travel up our noses they’ll bind to those receptor proteins, setting off an electrochemical signaling cascade that reaches your brain, and then your brain interprets that as a smell or maybe multiple smells.

Deboki: And here’s where we get back to dogs… First off, humans and dogs have a very long history of companionship. I mean, a human and dog have been found buried together from 14,000 years ago. Although it’s hard to find much evidence of this, because it was so long ago, there are historians who believe that early humans used dogs for hunting, helping us chase down prey but also detecting an animal that may be stalking us as their prey.

Sam: Dogs in the Middle Ages have been depicted wearing armor alongside knights and were thought to be used to track down missing people and even criminals. In WWII, mine detection dogs were used by multiple countries, and by the 1950s, formal K-9 training in the United States was taking off. Today, detection dogs are used throughout the country and they’re an important part of the Department of Homeland Security’s Chemical, Biological and Explosive Defense research and development.

Deboki: Sam and I wanted to understand: why dogs? So we called up Lucia Lazarowski, a professor at Auburn University in Alabama, and a lead researcher in their Canine Performance Sciences program. Her work is basically to make detection dogs better at what they do. So she and her colleagues may have a group or agency come to them saying they need dogs trained to find a specific odor or they need detection training in a certain kind of environment or scenario.

Sam: When Deboki and I got Lucia on the line, one of the first things we wanted to know was if a dog’s sense of smell is all it’s cracked up to be.

Sam to Lucia: I sort of always heard like dogs are a thousand times better at detecting odors than humans. And now there are people who are arguing actually, you know, humans are better at detecting scents than we thought before. But in the human-dog matchup, do you feel like dogs do have much better olfaction than humans?

Lucia: So I think that that comes from a lot of assumptions based on older research about humans not having a good sense of smell. And I think that there's newer evidence that humans have a better sense of smell than we're given credit for. That probably has to do with just different odors being more or less important for different species. And so we may specialize in odors that are important and relevant for humans that aren't for dogs and vice versa. But I still think that there is a reason that we utilize dogs for the detection tasks and not humans. We know from work that we've done and lots of other labs have done that dogs can detect diseases, they can detect tiny computer parts, trace amounts of explosives and chemicals. And just from working with these dogs every day we can walk into a room and see the dog immediately recognize an odor and find its location, whereas we don't notice anything.

And so I think in that sense a dog's sense of smell is much more sensitive than ours and they're capable of doing things that our noses can't do. And some of that might just be that there hasn't been a lot of research that has really put human olfaction to the test. And maybe with the right technology and tools and training that humans maybe could do some of these things more than we currently think we can. But from my experience and the current research that we have all evidence points to dogs being more capable of detecting odors than humans.

Deboki: So it’s not just a potential difference at the cellular level, it’s also the structure of a dog’s nose and nasal cavity that make it ideal for detecting odors.

Lucia: A lot of it comes to just the anatomy of the dog's nose. They have obviously longer snouts, and so they have structures in their nose that allows deeper and longer processing of odors than humans do because of our kind of shorter noses and the way our skull is shaped. They can actually smell with each nostril independently, so they can tell the direction of an odor based on which nostril is smelling an odor, whereas humans don't have that kind of bilateral sense. Dogs are able to inhale and sniff odors while they're also breathing and exhaling, whereas we can't do that. So they're much more efficient in how they can process odors.

Deboki to Lucia: So based on what you're talking about in terms of like the structure being so important to what they're being able to smell, do you find that there are certain breeds who are better able to do certain kind of smells or certain kind of tasks related to smell than others?

Lucia: So as far as their actual olfactory ability and sensitivity, it seems that there's not a whole lot of breed differences. Dogs as a species are very good. But when it comes to other aspects of what they're doing that influences their ability to detect, yes, there are differences. So brachycephalic breeds that have the short nose, they struggle to breathe so if they are searching a larger area, they're going to have trouble keeping up just with the physical stamina and being able to breathe efficiently while detecting. So in that sense, there's limitations to certain breeds.

Sam to Lucia: I have two Boston terrier/frenchie mixes, so they're little smushy faced dogs. They're probably never gonna be very good detector dogs, but if I was going to train them to be able to track down, you know, maybe track me down or even just track down a treat or something, could you give me a really basic step-by-step how I would start to, to train them to be able to do that?

Lucia: Yeah, there's actually a sort of infamous controversial study out there finding that pugs outperformed German shepherds on a scent detection task. So that probably had to do with other aspects of individual dogs and their motivation and behavior, but it did show that pugs are capable of doing these things. And it's pretty simple to train dogs to do these more simple odor discrimination type tasks because to them it's all a game. So it's all about making it a fun game to them where they learn that if I perform this task, I get a reward. So as far as teaching them to track a scent or to pick out a certain scent, you just wanna pair that odor with an award, like a treat or play or a toy. So you could start out with having someone hold your dog and then you walk, you know, a few footsteps away while dropping some treats and kind of dragging your feet in the grass and then the person can release the dog and they'll follow the treats and get to you and you make a big fun party out of it.

And then you can increase the distance of the path and the distance between the footsteps and kind of start fading out the treats until they're just following your scent alone.

Deboki: Sam did it work?

Sam: No, but really because my dogs are spoiled and know the cabinet where I keep the treats so they just kept eating treats and then running back to that cabinet. I’m gonna try again, but I’m expecting them to flunk out of detector dog homeschooling.

Deboki: Maybe not so shockingly I have never seen a frenchie as a detection dog, but I have seen certain breeds more often in certain contexts. For instance, I feel like I always see labrador retrievers in TSA lines, walking among people who are waiting to go through security. So I wanted to get an idea of why you’d use a lab versus, say, a corgi.

Lucia: It has a lot more to do with the environment the dog's gonna be working in and the specific nature of the task that we're asking them to do. So for example, if we're talking about explosive detection dogs there are some dogs that work in airports, for example, that are trained to detect explosive odors that are coming off of a person that's carrying that explosive. So kind of your suicide bomber or someone that's carrying explosive to go set down somewhere and then leave. So there are certain breeds that were bred originally for, for example, upland game hunting, where they're utilizing what's called air scenting, where they're actually sampling molecules in the air and following that odor to its source without really using a lot of visual targets. And they're able to detect an odor source and follow it in a big open area.

So the dogs that we tend to see doing a person screening tend to be dogs like pointers, Labrador retrievers. But you might also see beagles that are working the conveyor belts and screening the luggage and they're doing more of a sampling of individual packages that are coming through that also has to do with their size. So dogs that are working big areas and searching crowds and following people tend to be longer legged, larger breeds, whereas smaller breeds like beagles tend to work in a more confined area.

Sam: I’m never going to be in a TSA line again and not think about the odorant molecules being picked up inside these dogs’ noses.

Deboki: Yeah wow it’s amazing they can pick up such specific smells in a sea of people. Incredible. And also made me wonder if dogs are alone in this ability, so I asked Lucia.

Deboki to Lucia: Aside from dogs, do you know of other animals that are good at bomb detection?

Lucia: Yeah, so rats, actually, there's an organization called APOPO in Africa that I actually had the pleasure of visiting this summer. They're a nonprofit that does really incredible work training giant African pouched rats to detect landmines and demining areas that have mines so that people can live and work in these areas. And so the benefit there is that the rats are much more lightweight than a dog or a human. And so they can actually go into these areas and determine that the area's clear and safe or locate landmines where then we can come in and remove them. Elephants have a really great sense of smell, actually believed better than dogs. And there were some observations that elephants were just naturally avoiding areas that were found to have landmines. And so some follow-up research showed that they are also very capable of being trained to detect explosives and very good at it. Not quite as practical as a dog or a rat.

Sam to Lucia:
How, how would you train an, what, what is an African pouched rat? How would you train them?

Yeah, a giant African pouched rat. And actually the reason they're called pouched, people think that means that they have a kangaroo look, it actually refers to their cheek pouches.

Sam to Lucia: Oh, that's cute. I was picturing like a rat kangaroo combination.

They're probably the size of like a rabbit. Um, so they're not giant <laugh> in that sense, but larger than typical rats that we have here. But the training is the same basic principles as training a dog, so pairing the scent of an odor with a reward. They use things like mushed bananas and things that rats like to eat. Um, and just little by little the rats are trained to, to locate an odor, then they're given the reward and then the odor is buried a little bit further in the dirt and they have to kind of learn to detect the odor through interfering ground layers because what they're finding is landmines. So they need to find things that are buried underground. It's very similar to how we train dogs.

Sam to Lucia: That's really cool. Are they nice?

Lucia: They are, yeah. They act almost like little dogs. They're friendly. They follow their trainers around. They'll actually jump up on their shoulders and just kind of hang out and <laugh>, they're cute.

Sam: I had the cutest little ratatouille version of a rat in my mind when we were talking about this but then I looked them up and… ok they’re still rats. There’s something about that tail for me.

Deboki: Well Sam, if you’re looking for something that works kind of like a dog’s nose but isn’t a rat, there are detection tools out there and you’ve probably seen them used in airports or maybe entering concert venues. Anywhere that they’d want to be checking for possible explosives.

They’re called electronic sniffers or electronic noses. Just like a dog’s nose, they detect odor compounds floating in the air or from a swabbed sample. Either way, once in the machine the compounds are drawn through a glass tube that’s coated with molecules designed to interact with predetermined targets—let’s say an explosive like TNT—so if one of those compounds picked up by the electronic sniffer is TNT, it will bind to the molecule in the tube and cause it to fluoresce. Light sensors in the machine detect that fluorescence and tell you, “hey, TNT alert.”

Sam: So you might be wondering, if we have these, why still use dogs? Well there are a few reasons.

Lucia: The main advantage of a dog is that they are much more mobile and faster than these electronic devices. So a dog can quickly search a large area, quickly identify the source and location of an odor and, and go and respond to that odor,  clear large areas very rapidly. Whereas the electronic devices are meant for scanning small samples. So, for example, at the airport, a dog can screen hundreds of people or packages or luggage in a very short amount of time, whereas with the machine, you have to take a swab from the person's hands or from their suitcase and then process it through the device. On the other hand, dogs are living breathing beings, and so they need breaks, they need to rest, they need to use a bathroom, they can have good days, bad days, their motivation can fluctuate. But as far as efficiency, the dog definitely beats the machine. But the technology is advancing every day.

Sam to Lucia: At this point, dogs have shown to be very, very effective at detecting a huge range of odors. They're tried and true detectors of bombs, of landmines, things like that. I feel like there are so many reports of dogs being able to sniff out literally everything, you know, from cancer to… I don't doubt that they are remarkable, but I'm wondering in what field do you feel like we need a lot more information to be able to say, yes, dogs can definitely sniff out X, Y, Z. Do you feel like dogs really can sniff out Covid-19 or cancer or other medical conditions?

Lucia: As far as the research that we've done, we have not come across an odor that the dog couldn't detect. Anything that has an odor, which everything does, the dogs can detect it. It more comes down to how do we identify what are the particular molecules or markers that the dogs are picking up on. So all odors are very complex, so there's different things that make up what is being perceived. And so being able to identify what exactly in that odor is the dog picking up on, and how can we use that information to then be able to train them. Because really the challenge comes on the human side of when we are trying to train the dogs and present the odors we get in the way by incorrectly, assuming that the dog is picking up on a certain aspect of the odor when it's really something else.

Or we can inadvertently contaminate the odor in the way that we present it, or just by lack of proper controls, the dog may find that the way that we're presenting the odor, say in a certain packaging, they learn shortcuts where, oh, I'm just gonna learn to identify the smell of that package because we haven't controlled for that odor... As far as COVID-19, there's a handful of studies that have shown that dogs are capable of detecting it, but again, the challenges come in. How do you safely present the odor to the dogs, keeping the humans and the dogs safe? How do you make sure that they are actually identifying the odors associated with the virus? And then how can we actually implement this in a way that is effective in terms of protecting people from covid?

Deboki: Right this is such a good point. There are so many smells coming off of us. Our diet, our genetics… it all impacts how we smell and what a dog might be picking up on. Ok so, I think we should wrap things up with a story from Lucia that I thought was incredible, that shows just how deep an imprint that detection training makes on some dogs.

Sam: Yes, I loved this story.

Lucia: Some of the work that our program has done is with conservation detection dogs, where the dogs were trained to find pythons in the Everglades in Florida where the pythons were wreaking havoc on the environment. A lot of species were going endangered because of the pythons.

So we had our teams go in with dogs that were trained to find pythons and help researchers identify the prevalence of the pythons and the areas where they were and kind of help in that effort. So one of our dogs that was part of that research study where he was trained to find pythons in the Everglades after that project, he then came back and we've used him on other research projects and, this is maybe seven years after that project, he just went for a routine vet visit and started pulling his handler, like he smelled something and kind of dragged them through the vet clinic and then sat and responded on one of the staff in the vet office. She opened her lab coat and she had a pet python around her neck <laugh>. So even though he had not seen pythons, smelled pythons in years and had been trained to perform all other kinds of detection tasks since then, he immediately recognized that odor and remembered it after all that time.

Sam to Lucia: Seven years. That's amazing.

Yeah, we've had a lot of stories similar to that where dogs that were in our program and were trained on particular odors, they then get adopted out or they go on to have other jobs and do other things. They'll come across an odor from years before and immediately recognize it. And then we've done research, um, as well looking at dog odor memory and it does seem like odor memory, not just in dogs but in humans and other species too, is very resilient and very long-term once you learn an odor, you know, you can think back to scents from when you were in kindergarten or you know, you know, your grandparents' house. Odor memory is very strong.

Sam: The linking of odor to memory or… memory to odor? It’s so intense. Now we’ve done an episode on odor, and we did one on memory last year. Maybe we need an odor memory and nostalgia episode to really bring things together. Ideas for another day. Time for a Tiny Show & Tell.

If you want to kick off the tiny show and tell, go for it.

Yeah! Well, I want to talk about ice because I feel like ice is a thing that seems simple. It seems like, cool, it's got the water, it's got the H's, it's got the O's, and they make a nice little structure. It's simple, we all love it. But like everything, it's actually a lot more complicated. And I feel like this has been a thing that's coming up in things I've been reading lately. So I wanted to talk about it today in this particular news article from the New York Times about a new form of ice that scientists have made. So there are a lot of other forms of ice out there, and they're all basically the result of different combinations of temperature and pressure that change the way that the water molecules are arranged together within ice.

So there are around 20 crystalin variations of water that we know of right now, but there are also two amorphous forms of ice, which basically means that they're not the nice orderly crystalline structure that we associate with the water molecules in our favorite ice. It's a lot more jumbled inside of there. And so you can have high density ice, which is more dense than water, or you can have low density ice. And these types of ice are usually found in space like in comets or on icy moons like Europa. And researchers managed to create another form of amorphous ice, and they managed to make medium density ice, which I love. They've found the range.

So the way they did this is they started by chilling ice to minus 320 degrees Fahrenheit. So they just took some liquid nitrogen, chilled it down a lot, and then they put it into a container with some steel balls and then used that to basically shake the ice a lot. So as they're shaking the container, the balls are shaking into the ice and breaking it all into smaller pieces. And when they opened that container up, they found that the result was that medium-density ice. And basically, somehow the force of those steel balls on the ice crystals was enough to knock the water molecules out of their crystal positions and into this more amorphous structure.

And one of the things that's cool that I didn't think about is basically this might mean that this medium density ice, where they called it MDA, is like a glass form of water because glass is essentially a liquid that was flowing until it cools down, but freezes without actually crystallizing. And so that's kind of what could be happening here. And it's also interesting, one of the things that I did not appreciate until I started for some reason having ice come up in my reading a bunch this past week is that learning about these different forms of water teaches us more about what space looks like because this medium density ice could be what, like I said earlier, places like Europa might look like in terms of their water. So I just thought that was really cool.

Sam: I love that. Medium density ice. I understand conceptually, but I'm like, how do you even go about designing an experiment to create "medium density ice"?

Deboki: Yeah.

Sam: It's just kind of mind blowing to me. But I'm so glad that there are people whose brains work this way so they can design the experiments and not me.

Deboki: Yeah, I agree.

So my tiny show and tell is not about water, but it's about things that live in water. It's about fish. So I'm making a connection here. All right. So hundreds of thousands of years ago, Mexican tetra river fish found themselves trapped in the caves of one of Mexico's rainforests. There was no way out. They had two options. Adapt or die. And fortunately, over thousands of years, they did adapt, and that has meant massive changes in how they look and how they behave. So not only are they blind and a lot paler than their ancestors, they've also had to become total couch potatoes to make sure that they're not wasting their energy in these dark dismal caves that have very little food and nutrients. So they're just conserving energy. That seems to have led them to store a lot more fat.

In humans, we know that if you're being super sedentary, you're not moving around, you're losing muscle mass, it's not great for your health. It's linked to things like heart disease and stroke. But in these fish, it turned out that they're still really healthy. So researchers have figured out that they've actually undergone genetic changes that have altered their muscle metabolism. So they're able to actually turn this fat that's stored in their muscles into energy super fast, and it allows them to actually swim just as quickly as their surface level ancestors if they need to, which I think is really interesting, just this adaptation that allows them to not only continue to be healthy, but also in a pinch, essentially reactivate their muscle strength that they're not using on a day-to-day, which is cool.

So unfortunately, we humans don't have this couch potato superpower. We still have to move around, get some exercise to stay healthy. I found that this was sort of an interesting mental exercise type of paper that came out where it makes you wonder thousands of years from now if we continue with these more sedentary lifestyles overall, will we also have adapted in some way to be slightly healthier couch potatoes? I don't know. That evolution is not going to happen within a generation. So this is not like a PSA for people to just chill out and see what happens.

Really, the researchers think that understanding these genetic changes and how different molecular pathways have been altered over the course of evolution, maybe it could help us understand what to target or look for if in the future we want to do things like take super long space flights to other planets or galaxies that last years and years. We'll be stuck in a ship, we can't move around very much. What can we learn from cave fish and these really cool adaptations that maybe could be used to protect us from disease that comes with being really sedentary. So I just thought it was kind of interesting. It's like a very cool evolutionary development paper, but at the same time, it gives us an opportunity to just think about the possibilities of what you could learn and why it might be applicable to humans.

Deboki: I love the idea of that application. And then again, I think it also funnily ties back to my tiny show and tell where the water is also about space. Initially when you were describing applications, I was trying to guess what the possible application would be besides me getting to fully indulge in my love of sitting around all day? But it's like, oh, right. If we want to make going to space a sustainable long-term thing that people could do, we've got to figure out how to get away with not moving for very, very long periods of time.

Sam: Maybe this could be the beginnings of a movie about space travel, and it's only possible because we learned from cave fish that X, Y, Z. It just made me think this could be the beginnings of a blockbuster about space travel to other galaxies.

Deboki: Totally.

Thanks for tuning in to this week’s episode of Tiny Matters, a production of the American Chemical Society. Our exec producer is my co-host, Sam Jones.

Sam: This week’s script was written by me and was edited by Deboki and by Michael David. It was fact-checked by Michelle Boucher. The Tiny Matters theme and episode sound design are by Michael Simonelli and the Charts & Leisure team. Our artwork was created by Derek Bressler.

Deboki: Thanks so much to Lucia Lazarowski for joining us. If you have thoughts, questions, ideas about future Tiny Matters episodes, send us an email at You can find me on Twitter at okidoki_boki.

Sam: And you can find me at samjscience. See you next time.