You're lost in the wilderness. Now what?
24:04 Share

Silicon Valley Bank is still the SVB you know and trust. The SVB that delivers human-focused, specialized lending and financial solutions to their clients. The SVB that can help take you from startup to scale-up. The SVB that can help your runways lead to liftoff. The only difference? Silicon Valley Bank is now backed by the strength and stability of First Citizens Bank. Yes, SVB. Learn more at svb.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.com.

Think scaling AI is hard? Think again. With Watson X, you can deploy AI across any environment. Above the clouds, helping pilots navigate flights. And on lots of clouds, helping employees automate tasks. On-prem, so designers can access proprietary data. And on the edge, so remote bank tellers can assist customers. Watson X works anywhere, so you can scale AI everywhere.

Learn more at ibm.com slash Watson X. IBM, let's create. Paul Daugherty always knew he wanted to go out West. So when he got the opportunity to work as a park ranger in Yosemite, it was a no brainer.

But he didn't realize just how much time he'd spend looking for stranded hikers. You know, growing up in the Bronx, I didn't really know much about search and rescue. I thought that's just what firefighters did when there was a burning building. And I basically went to like the mecca of search and rescue. Yosemite's a huge park, almost the size of Rhode Island. And people get lost there a lot, more than in almost any other national park in the U.S.,

It's wild, isolated terrain. And walking a few minutes down a trail can leave you without cell service and far away from anybody else. Which is exactly what happened on one of Paul's first search operations. It was 2008 at Washburn Point. This high overlook where you can see these tall waterfalls cascading over giant rock formations. A hiker wanted to spend some time alone, so his friends left him there while they explored the park. And when they came back, he wasn't there.

And then they eventually called 911, thinking maybe he had hitchhiked back home or he could be lost in the park. You know, all these different scenarios. So the ranger started collecting clues. They asked his friends when and where they'd last seen him and any other personal details that might help with the search. They said he played soccer, had no known health issues. He didn't necessarily have water, right? He definitely didn't have food. He didn't have overnight clothing. So...

We sort of knew in that scenario, there's a missing person, they could probably cover a lot of ground, and they don't necessarily have all the equipment to make it in the backcountry. The clock was ticking. This was a life-or-death situation. So the rangers got to work pulling out paper maps and circling potential search areas around Washburn Point. At first, the rangers said, can you just print out more maps?

Paul was surprised because he'd studied how to make complex digital maps, and he was already using them to track crime in the park. What the rangers were doing seemed so old school.

So we offered to make them a digital map instead. Then they said, you can do whatever you want as long as you get us that paper map because we, you know, we drew all over it and we need another one. So Paul made a digital map and plotted out where to send search teams to cover the most ground. After a few days, with the help of almost 200 volunteers, a park helicopter eventually spotted the lost hiker on a completely different mountain. We did find them alive. And so that kind of set the bar for at least my team. We're going to do this every time. And

And that's when I started reaching out, like, can I use all the science I could? Those other rangers were confident in how to search for lost people. But without Paul's digital maps, the search might not have been as successful. Paul saw how the rangers had a major blind spot, and that made him wonder about his own. Do we really understand how missing people move in the environment, right? For all the different scenarios that can happen, where are we most likely to find them?

This whole problem of search and rescue seems like it should have been solved already. We have phones and GPS and so many ways to keep track of each other. But it's a lot easier than you'd think it is to get lost and a lot harder than you'd think it is to be found. I'm Manning Nguyen, and this week on Unexplainable, how do we find someone when they get lost? ♪

When someone goes missing, rangers like Paul structure their search around two important concepts. Probability of detection and probability of area. So instead of a missing person, let's just say you lose your keys in your house.

Probability of detection is about how good of a job you're doing looking for your keys. How likely am I to detect the thing I'm looking for, right? So maybe I'm looking in my room, but I'm not looking under things. Or some people have a way to, like, locate their keys with their phone, right? Am I using all the tools and the right tools needed to do that search and detect the object that I'm looking for? And then there's probability of area. Okay.

Am I looking in the right room? Are the keys in this house, right? If I don't have that right, it doesn't really matter what I do next, right? If I'm not looking in the right place for those keys. When it comes to finding lost people, we've gotten a lot better at the probability of detection. We have things like drones, thermal mapping, and GPS to pinpoint people. There's even research on how to train search dogs to track down smells better.

But knowing where to deploy these tools is its own stubborn problem. A lost person can make any number of choices, interact with the environment in so many different ways. So how do search teams know where to look? They might vote on the most likely areas the person could be or comb popular trails.

But one of the most important tactics for finding a lost person is looking for clues in the past. We want to look at historic data, right? Where does the data tell us to search, right, or guide us to search? And that's what that Lost Person Behavior book by Bob Kester provides. How can you use past cases to predict where the missing person's going to be?

Robert Kester has been in search and rescue for over 30 years, and he's most famous for a book called Lost Person Behavior.

The Lost Person Behavior Book is a search tool, a field guide. But really, it's the front end of a database. It's the result of a decades-long effort to create and analyze a huge volume of lost person cases. It's columns and rows and columns and rows of data. No other lost person database comes anywhere close to its scale. Roberts recorded over 300,000 cases and counting from all over the world.

And he's still not done. I think I identified my end point is when I have about 4 million cases, I'll probably be satisfied. And I'm still a long ways from there. If you flip through the Lost Person Behavior book, you'll find chapters for different kinds of people who get lost. So a hunter is going to behave differently than a hiker who's going to behave differently than a mushroom picker.

There's 41 categories of people in his book, and each one comes with a sort of archetype of how different kinds of people move. It comes with a checklist of steps to start looking for them, and it tells you what clues to look for. So, for example, I like mushroom foraging. That's on page 178.

Robert's book says, "I'm highly secretive about the areas I visit to keep my mushrooms hidden. I'll often go off trail, I'll jump from spot to spot without paying attention to the general area, and if I get lost, it's most likely because I have poor navigational skills." Okay, that's all pretty true. But the book is also full of statistics, like the chances a lost person is still alive after a certain number of hours, or how often they're found near something like a river.

And the most useful details here are the average distances between where the person was last known to be and where they're found. It's that probability of area I mentioned before. And these numbers can be pretty different depending on who you're trying to find. So 75% of the time, a mushroom picker like me is found within four miles of their last known location. For a climber, it's two miles. For someone with Alzheimer's, about one. By using the probabilities,

you can create a heat map and say, "Here are more likely areas and here are less likely areas." If you were to draw the probabilities on a map, you'd mark the point where the person was last seen and then draw four concentric circles around that point.

So the smallest ring would represent the area where 25% of people were found in this database. Then you draw bigger and bigger rings until you have something like a bullseye on your map, helping you figure out where to concentrate your search. It's going to be useful about 95% of the time. Beyond that, everything else is a statistical outlier. And it's darn hard to plan for the statistical outliers.

Robert's book is used in search and rescue trainings all over the country. And every single one of the search and rescue volunteers I spoke to, from Alaska to Virginia to British Columbia, said they find it useful. So did Paul. Going all the way back to when I first started doing search and rescue, I was looking for cookbooks, right? Like, what's the standard operating procedure for a missing person? Like, how do I do this thing, preferably with a checklist because I'm stressed and

I thought it was really useful. It gave us guidance on roughly as the crow flies, how far away are people typically found based on different scenarios. So it

it provided the ability to take action, right? And that was really, really good. Paul used this book on a lot of his searches. But after a few months, he started feeling like it had some major limitations. You're looking at this giant map of Yosemite. Where do we start, right? And that book, that was my only kind of guide, right? And so I kind of found it useful to get started there.

but found it limiting in that we're overestimating this search area and we don't have the resources to search it. And that was really what got me curious was, you know, we've got this whole park to work with, maybe even beyond. How do we focus our resources? We have this lost person behavior book, but have we really taken the science as far as it could go? That's coming up after the break.

Your business deploys AI pilots everywhere. But are they going anywhere? Or are they stuck in silos, exhausting resources, unable to scale? Maybe you don't need hundreds of AI pilots. You need a holistic strategy.

IBM has 65,000 consultants with Gen AI expertise who can help you design, integrate, and optimize AI solutions. So you're not just deploying AI, you're scaling it across your business. Learn more at ibm.com slash consulting. IBM, let's create.

Thank you.

It's time to cut through the noise and make a real impact. So tune into the future of marketing, a special series from the PropG podcast sponsored by Canva. You can find it on the PropG feed wherever you get your podcasts.

Creativity is one of the core traits that makes us human. It allows us to tell stories, to create, and to solve problems in new and exciting ways. So why does it feel so threatened? With new technological advances that can create art in milliseconds, where does that leave us? In this special three-part series, we wanted to ask, how can we save and celebrate creativity? To

Tune into Saving Creativity, a special series from the Gray Area sponsored by Canva. You can find it on the Gray Area feed wherever you get your podcasts.

When Robert started the Lost Person Behavior Database, he had to figure out how to analyze and parse all this information he'd collected from various federal departments, state agencies, local volunteer teams, police departments, basically anywhere he could get it. The hope was somewhat biased to basically help myself. What information would help me run a search? And

And I felt like as long as I kept it focused on what's useful to practitioners, then in the end, not only was it going to help myself, but it was going to help other search planners. There's a million ways to analyze a dataset and parse out patterns and probabilities. Robert decided to take a more human-centered approach. He landed on 41 categories for people, but only a handful for terrain and regional differences. And that's what you see in his book.

And Paul wanted to dig deeper, but the database that Robert made was private. It was proprietary. At the time, all you could do was get the book. He didn't have the raw data. All he had were summary statistics, like how far, on average, did a lost mushroom forager travel in a place like Yosemite that's classified as a temperate region.

But the coast of Spain is also temperate, and so are the Great Plains. And they don't feel at all like Yosemite. We have this unique terrain environment. If you go missing in Yosemite Valley, you've got like a 3,000 to 4,000-foot toilet bowl to climb yourself out of, right? You've got this big, huge terrain challenge. And even with great trails, that's going to slow you down. It's probably also going to affect your intended destination. And I just didn't trust that a summary that was kind of

of the whole world was going to be able to tell me everything I needed to know about the area that I was working in. For any search, there's a limited number of people and tools and resources to rely on. So if search teams were just using the Lost Person Behavior Book, there was a chance they might have been stretching those resources too thin, maybe to a breaking point, with the lost person's life on the line.

So Paul started plotting out a study to compare the profiles in the book with Yosemite lost person data. But to do so, he needed to build his own database. Because, incredibly, Yosemite, which has more search and rescue incidents than almost any other park in the U.S., had never digitized its lost person files. There were paper reports. So I had to go and get them, dust them off, read them, and glean whatever information I could.

Paul spent hours in the archives sifting through these reports. He was looking for specific locations and coordinates, but he also found a lot of detail about how the searches were done. You can sort of see

Paul recruited a small team of people, and they managed to clean up thousands of lost person reports. The data confirmed his suspicions.

that people in Yosemite ended up in significantly different places than the Lost Person Behavior book said they would, which meant wasted time and manpower spent looking in the wrong areas. So Paul and Robert worked with some collaborators to chisel those oversized probability rings to better match the contours of the landscape. Before this research, rescuers looking for a hiker might have used the book to draw the same bullseye rings onto a map of Yosemite that they would have drawn in the Rocky Mountains or the Appalachian Mountains.

But now, they could sculpt those rings into a more customized shape, which means faster search operations and fewer injured or dead people. There's still a lot more to do. The science of search needs to keep evolving. On one hand, Robert's continuing to refine his own database, making more categories and doing more complex analyses of it. He's also giving the data to more researchers who want to compare it to models of their own. On the other hand, a lot of this science depends on having good data in the first place.

There's no centralized search and rescue agency in the U.S. that keeps track of lost people.

It's local teams, mostly based on volunteers doing these searches in their spare time, who are at the forefront of this work. Even if they only have one or two missing person cases a year, and maybe in 10 years they've got a sample size of 20 or 25, those are stories, right? There's a ton of information beyond the math data that can help them. Like this one story Robert told me. A long time ago, he did a search for a father and his son.

They took a wrong turn on the Appalachian Trail and ended up in a drainage ditch. No idea where they were. They scrambled up the mountain, reached a ridgeline, and then kept going, getting more and more lost. They ended up camping two nights, going back and forth and up and down, before the father was finally found with his son near a river. If you actually just looked at where he was last seen and where he was found, you would never figure out any of that.

Because the distance between those two points was about three quarters of a mile. The distance he actually traveled going up and then back down, and if had been given the chance, he would have gone back up again, was about 15 miles. From where they started to where they were found, it looked like they barely moved in three days. But the story can tell you where they made the wrong turns and what parts of the terrain were confusing to them.

And that knowledge can sometimes help teams find lost people faster, which is important because the longer a person's lost, the more likely it is that they'll never make it out. And that's not something that raw data alone, no matter how well you structure it, can really do. Another search volunteer I talked to told me about a team in Canada that used lost person behavior stories to get a really good sense of where lost skiers ended up, depending on which trail they started on.

So skiers who turned off the, let's say the blue trail, ended up in the western ravine, while skiers who got lost on the green trail mostly ended up in the eastern one. This key insight totally transformed how the team led searches. Not only could they make searching the right ravine a priority, they also installed helipads and cabins in those ravines so that the skiers could hunker down and wait to be rescued instead of wandering deeper into the wilderness.

When someone gets lost, they leave behind so many questions and uncertainties. And a search planner has a lot of tools in her belt to find them. There's statistics, stories, the investigative information. But the challenge is to figure out when to rely on which tool. Because there's no silver bullet here. There's only people like Paul, trying their best with what they've got. As I worked on more missing person cases, you meet the family, right? Showing them on a map what has been searched,

Getting lost is scary and lonely.

But there's something weirdly comforting about the fact that if I ever get lost much from foraging somewhere, it's probably not the first time that someone's gotten lost there and in the same way. And it makes me feel a little less alone to know that there will be a whole team of people looking for me, that they've learned from the thousands who have gotten lost before me, and they're using all the tools they have to bring me home. This episode was reported and produced by me, Manning Nguyen.

And I just want to say one more thing. If you're going out, even with another person, tell someone where you're going and how long you'll be gone. It's the best thing you can do to keep from becoming a database entry yourself. We had editing from Jorge Just and Meredith Hodnot. Sound design and mixing from Christian Ayala. Music from Noam Hassenfeld. Fact-checking from Anouk Dussault. And Bird Pinkerton looked up. She saw an enormous domed ceiling engraved with a tortoiseshell pattern.

Giant 60-foot pillars span the height of a great hall. She had made it to Tortopolis. Special thanks to Michael Coyle, Helena Michalis, Sally Dickinson, Cassandra Aguirre, and Evie. If you have any thoughts about the show, send us an email. We're at unexplainable at vox.com, and we'd love to hear your thoughts, your criticisms, and your suggestions. And if you can, go leave us a review or a rating wherever you listen. It really helps us find new listeners.

This podcast and all of Vox is free, in part because of the gifts from our readers and listeners. You can go to vox.com slash give to give today. Unexplainable is part of the Vox Media Podcast Network. We're off next week, but we'll be back in your feed in September. See you later.

Canva presents a work love story like no other. Meet Productivity. She's all business. The Canva doc is done. Creativity is more of a free thinker. Whiteboard brainstorm. They're worlds apart, but sometimes opposites attract. Thanks to Canva.

The data is in the deck. And now it's an animated graph. Canva, where productivity meets creativity. Now showing on computer screens everywhere. Love your work at Canva.com. And we're back with Canva Presents Secret Sounds, Work Edition. Caller, guess this sound. So close. That's actually publishing a website with Canva Docs. Next caller.

Definitely a mouse click. Nice try. It was sorting 100 sticky notes with a Canva whiteboard. We also would have accepted resizing a Canva video into 10 different sizes. What? No way. Yes way. One click can go a long way. Love your work at Canva.com.