Viterbi students program underwater gliders

Doctoral student researchers in several fields are collaborating to explore the possibilities of engineering, computer science and biology through the use of autonomous underwater vehicles.

Students in the Viterbi School of Engineering’s Robotic Embedded Systems Laboratory and USC(sea) Lab have been able to do this by programming underwater gliders that will roam the sea, testing and gathering data that is essential to biological research conducted by other students.

Chairman of the Computer Science Department Gaurav Sukhatme, who is overseeing the project, said the research is actually composed of a series of projects.

“We are interested in understanding how to build robots that are intelligent and that can survey the ocean,” Sukhatme said. “This has been about a decade-long activity, so what you are seeing are our latest set of experiments in our long-term endeavor.”

Sukhatme said the project is not without challenges, especially when it comes to adapting computer science for the rigors of real-world applications.

“We are a bunch of computer scientists, and we are specializing in computer intelligence,” Sukhatme said. “But just to work in the actual ocean, we have to do a lot of things to integrate our ideas with ocean-going hardware and get it to work reliably in the ocean.”

Arvind Pereira, a fifth-year doctoral candidate studying computer science and the student leader of the lab conducting the joint research effort, said the research group is augmenting pre-existing technology with new technology to fit the project.

“We decided not to make our own glider because we did not want to spend time developing a system that is already out there and seems to be working quite well,” Pereira said. “Instead, we wanted to focus more on engineering the new things in it, which would allow us to use the preexisting glider and make it smarter. In computer science we do that through improvements in software and hardware.”

Dave Caron, a professor of biological sciences, said these vehicles are not only cost- and time-efficient, but can also stay out on missions for a month at a time and reach places where human divers could not go.

“The idea of the robotics is that we will have a continuous presence in the ocean and that is basically what we have developed,” Caron said. “We have developed technology and approaches that allow us to know what is going on in the water on a constant basis, so that when something is beginning to happen, we can be there as it unfolds, rather than be there after the fact. That way we try to understand what the environmental conditions are that give rise to the toxic algae that we see in the water.”

Caron said the collaboration and teamwork better these very different fields is necessary to achieve both groups’ ultimate goals.

“It’s a huge environment, so if we tried to keep track of things by simply going out there, we would not be able to cover the area or the time,” Caron said. “We want to be able to have a finger on the pulse — know exactly what’s going on and when things look like they will be informative to us. Then we can get out on the ship and do our sampling and experimental work.”

Caron said he works closely with the bioscience side of the project, which has been going on for almost a decade. During this time, he has seen results in not only the project itself, but in the students participating as well.

“I think the students benefit tremendously because they get essentially an interdisciplinary training,” Caron said. “They not only are versed in the robotics and the computer science of it, but they understand … how their approaches can benefit other aspects of science and engineering and, in this case, the biology of coastal oceans.”

Bridget Seegers, a doctoral student studying biology, said the intertwining of science and engineering has been mutually beneficial for both sides.

“[The robotics group] definitely has a different set of skills than we have, so it’s been nice having them help us with data processing. They do things a little more efficiently that help save us some time,” Seegers said. “Even though you are working together, you have slightly different goals, because we are interested in the science and they are interested in the engineering aspect. I think we have come to the spot where we can do both those things. There is a lot of understanding between the groups and working together to make sure everyone’s goals are realized.”

However, Caron said collaboration and cooperation is not always easy. With so many different ideas and backgrounds, the diversity presented its own problems.

“The most challenging aspect of this has been that we have biologists and engineers and computer scientists and robotics people sitting down at the same table and trying to talk, and we often have very different jargon, very different terminology,” Caron said. “We spend a great deal of time making sure we understand each other when we start speaking in our respective jargons. The communication issue is a significant one, and we do spend a lot of time on that — making sure that we really understand each other when we talk.”

Caron also noted other challenges the team faces.

“The other thing that I think is challenging and that we work very hard at is that we always try and pick a topic that will be interesting to both the engineers and the biologists and to the computer scientists in general,” Caron said.

Though people focus on their individual interests, Caron said it benefits the project’s progress.

“We have people who want to do just the hardware robotics, we have people who want to do the computer programming, we have people who want to use the information to understand biology and we try to make sure that whatever problem or issue that we pick is interesting in all aspects of those endeavors so that everyone has research to do.”

Overall, Sukhatme praised the effects collaboration praised the positive effects of collaboration.

“It’s wonderful,” Sukhatme said. “It’s absolutely superb. We learn a lot from each other and it makes us ask questions that we wouldn’t normally ask.”