That's part of the answer that Professor Christopher Lowe gave me when I asked him about why he needs a shark tracking robot. Lowe heads the Sharklab at California State University Long Beach. The goal of the lab is to learn more about shark behavior - why they do the things they do and in what context they do them.
The lab's goals were made a little easier to achieve starting four years ago, when Lowe was contacted by Professor Christopher Clark from Harvey Mudd College. There, Clark directs the Lab for Autonomous and Intelligent Robotics, which focuses on applications for multi-robot systems. Clark contacted Lowe to gauge his interest in building an autonomous robot that track sharks in ways that humans simply can't.
After receiving a grant from the NSF, the research team began work on developing a robotic shark tracking system. The primary robot is off-the-shelf - they use Ocean Server's Iver2 autonomous underwater vehicle. But they've added their own custom modifications.
For starters, the team added two hydrophones to each side of an Iver2. The hydrophones receive a signal from the radio tag on the shark that the team is interested in tracking. By using two of them, the computer on the robot is able to ascertain the bearing of the shark they're looking for, and steer the robot accordingly. Then, for the programming of the propulsion systems, they "trained" the robot the same way you'd train a shark tracking grad student.
"Students have a tendency to chase the sharks and get too close," Lowe told me. "You want to get close to get an accurate position, but if you get too close, the shark's behavior changes. So we have to train students to find the right spot."
The robots work the same way. Out in the ocean, the team actually uses two of the robots, each of which are working to track a given shark. The two robots communicate with each other to control their movements.
"What we've done is created a circle," Clark told me. "The AUV's are located on opposite sides of the circle, keeping their distance from the shark. But because they're on different sides, it's easier for them to catch up if a shark suddenly moves in one direction."
In addition to tracking sharks, the robots are also equipped with a number of sensors. These measure things like water temperature, salinity, pressure, and other factors that give biologists more insight to how sharks behave.
"They build a 3-dimensional world that the shark is swimming through," said Lowe. "It tells us where we are in that world, and helps me understand how animals make decisions."
And unlike people, the robots can sample the water and track the shark at the same time. That's something that a person can't do. The robots can also get more data about where the shark is at a given time. It picks up a signal from the tagged shark every 2 seconds. People are sometimes able to get a more accurate position, but the data points are more spread out through time.
Overall, the work that the team is seeing from the robots will provide opportunities to learn things about sharks they never have before. In a few weeks, they plan to do a study of the behavior of baby Great White sharks - something no other team has done yet.
"For me, it's revolutionary," said Lowe. "It changes the way we do this kind of science."
