PROVO — When the trail is gone and the scenery doesn't look familiar anymore, some hikers or hunters begin to panic. They may go downhill, follow other paths or drainages or simply wander in random directions — making them more difficult for search and rescue teams to find.
BYU Ph.D. candidate Lanny Lin and professor Michael Goodrich are studying those seemingly random behaviors as well as using complicated mathematical models and topographical maps to come up with another tool for rescue teams.
Their goal is to predict the likely locations of a victim, based on the foliage and terrain, GPS records from other lost hikers and advice from search and rescue experts.
"This very powerful mathematical tool (is being used to solve) an interesting, socially relevant problem," Goodrich said. "How do people behave if they're lost and they're walking through wilderness that has natural challenges?"
The goal, a year or two down the road, is to have a computer program where an incident commander could download the area's topographical map, type in the last known location of the victim and extrapolate out.
"It's kind of like a ripple effect," Lin explains. "By combining all these different ... probabilities, you can say there's a 20 percent chance the person would travel this certain direction, then another 30 percent chance they would travel another direction. If the person had been reported missing for three hours, you can run this model, predict three hours from the last point seen and it will give you an idea of some of the high priority places you probably should be looking."
The mathematical model itself is not new, but blending it with behavioral models is what won the admiration of the journal Computational and Mathematical Organization Theory, where Lin's and Goodrich's article was recently published.
"These models are only going to be one tool, and they aren't intended to substitute for the judgment of these search and rescue experts," Goodrich said. "Sometimes they have intuition and hunches that they need to follow."
These new modeling efforts are also being combined with BYU's unmanned aerial vehicles, with their GPS systems and video and infrared filming capabilities.
The hand-launchable UAVs are a safer and cheaper searching option than manned aircrafts, but in order for rescuers to know where to send them intelligently, they need a list of most-likely places, Lin said.
Goodrich said they're in the process of talking with the FAA to learn about the rules and to get permission to fly their planes during actual rescues.
Ron Zeeman has been on the Utah County Sheriff's Search and Rescue team for 10 years and knows that victims don't always fall into predictable movement patterns.
"If Lanny's program can help ... guide people to where to start looking, then it's great," said Zeeman, who also works at BYU's OIT Operations Center. It was Zeeman who was instrumental in getting the planes geared up for search and rescue operations in the first place.
"We've been trying really hard to focus on problems that are very highly socially relevant," Goodrich said. "Not very many people are looking at technologically-based solutions to help search and rescue."
e-mail: sisraelsen@desnews.com