News Photo by Zipporah Abarca Katie Skinner (far left) and other members of the multi-institutional collaborative research team; Mason Pesson (middle left), Will Ard (middle right) and Corina Barbalata (far right) pose with some of the ROVs they’ve been deploying for AI sonar-data collection.

For the past two years, Katie Skinner and a multi-institutional collaborative research team have worked on developing new technology to collect data and make discoveries in the ocean and the Great Lakes.

Tonight Skinner will present the work that has been done with the project thus far. The event is part of the Thunder Bay National Marine Sanctuary Lecture Series. The discussion takes place at 7 p.m. at the Great Lakes Maritime Heritage Center and admission is free.

Skinner is an assistant professor in the department of robotics at the University of Michigan. Originally from Long Island, New York, she completed her bachelor’s of science in mechanical and aerospace engineering at Princeton University. Then her journey to Michigan consisted of attaining her masters and Ph.D. at the University of Michigan in robotics.

The research team she has been working alongside for the past two years includes Corina Barbalata from Louisiana State University and two graduate students, Mason Pesson and Will Ard. Skinner said there are also grad students participating from the University of Michigan and Michigan Technological University for the Great Lakes Research Center.

“It’s been really great,” she said. “Working together as a team, we all bring different expertise.”

Skinner said the Michigan Tech team has worked on deploying its autonomous underwater vehicle to do sonar surveys of Thunder Bay and Lake Huron.

According to L3Harris Technologies, these AUV’s are the first commercially developed, low-cost sea vehicle. They specialize in environmental monitoring, sensor development and sub-surface security.

Those working at the University of Michigan have put together computer system methods for automatically processing the sonar data collected from the large area searches, Skinner said. This is how sites of potential interest are detected.

“We have come to Thunder Bay, because there’s a large amount of known shipwreck sites that we can survey to really test our methods,” Skinner said. “So that’s really exciting for us to be here.”

Skinner mentioned how it feels to be able to watch these systems in action as they are deployed and solve different technical challenges. She said it motivates her to come up with new solutions to any remaining challenges.

“… You can spend a lot of time in the lab at your computer, coming up with solutions,” Skinner said. “But whenever you go out into the real world, you’re going to find new problems that you’re going to have to overcome.

So that’s really exciting, because it lets you really employ problem solving and creativity in action, when you’re trying to get through those challenges.”

The project is funded by the National Oceanic and Atmospheric Administration Ocean Exploration program. The research team has worked with NOAA sanctuary scientists in Thunder Bay as well as the state and conducted expeditions in the Thunder Bay National Marine Sanctuary.

Skinner’s discussion will include how the group has developed technology for robotics and artificial intelligence to help accelerate ocean discoveries of shipwrecks and other sites that may be of interest.

She said there will be an overview of the robotic systems the research team has been using for deployments as well as the sensors. Additionally, there are new methods the team has conducted for AI to make the robots smarter that she will present on.

These new methods include processing side-scan sonar data to detect shipwreck sites without it having ever seen a real shipwreck site before.

According to NOAA, side-scanners are used for finding and imaging objects on the seafloor.

“The main challenge really is that we have relatively limited data of shipwreck sites compared to other applications; we might think about deploying AI algorithms like autonomous driving applications, for example,” Skinner said.

Our goal is to be able to develop methods that are still able to learn how to detect sites of interest without having seen, you know, tens of thousands of sites. Over the last two years, we’ve collected a little over 20 distinct sites of the known shipwrecks that are here in Thunder Bay.”

The research team has not found any new shipwreck sites in Thunder Bay, however Skinner said it has detected some interesting drop offs in the lake.

Everytime the team goes out, the potential of discovering new things is always present, Skinner said.

“It’s just really exciting, that we’re really exploring the world around us,” she said. “Working in Thunder Bay, it’s been really incredible to be able to work alongside marine archaeologists here, who dive the sites regularly.

“They really know them in and out and it’s really exciting to be able to see it through their eyes and to start to understand the importance of preserving these culturally significant assets.”

The past two years have consisted of being able to observe the importance of science and the development of technology, she said.

In her personal experience, Skinner said she has found there is a lot that can be done to develop new technology for ocean exploration.

“Every time we go out into the oceans, or the Great Lakes, there’s just so much to see and so much that we don’t know that we can still uncover, so it’s (a) really exciting area to work in,” she said. “And there’s still so much to do in terms of how we can deploy new vehicles and new technology to explore these areas, and also to make these systems more intelligent, so that we can do that more efficiently.”

After the event on Thursday, Skinner said she hopes the audience can see how robots and AI can be utilized to explore the world and make new scientific discoveries.

This story was produced as part of the Michigan News Group Internship. Zipporah Abarca is working for WCMU this summer at The Alpena News.