Algorithms Devised to Aid Inspections of Mines on Ship Hulls

Hovering Autonomous Underwater Vehicle (Bluefin Robotics)

Engineers at Massachusetts Institute of Technology have written algorithms that help underwater robots inspect the hulls of ships for mines as small as an iPod. The findings of mechanical engineering professor Franz Hover and Ph.D. candidate Brendan Englot will appear in a future issue of International Journal of Robotics Research.

The algorithms help guide a robot called the Hovering Autonomous Underwater Vehicle (HAUV, pictured left), originally developed as part of MIT’s Sea Grant program, that uses a sonar camera to emit signals and process the pings coming back. Researchers in MIT’s Autonomous Underwater Vehicle Lab founded Bluefin Robotics, a company in Quincy, Mass. that is commercializing the HAUV technology, and with whom Hover and Englot collaborated on the project.

The goal of the algorithms is to enable the device to produce a resolution fine enough to detect a 10-centimeter mine attached to the side of a ship. As Hover explains, “A mine this small may not sink the vessel or cause loss of life, but if it bends the shaft, or damages the bearing, you still have a big problem.”

The problem involves protecting a huge target — a large commercial or naval vessel — while searching for a small feature. Hover and colleagues took a two-step approach in solving the problem. The HAUV first approached the ship’s hull from a safe 10-meter distance, navigating in a square around the vessel. The researchers processed the the returned sonar signals into a misty point cloud that showed larger structures like the ship’s propeller, but not items as small as a 10-centimeter mine.

Using the point cloud, the researchers processed their sonar data with computer-graphics algorithms to generate a three-dimensional mesh model that identifies the ships major structures. This mesh model enabled the HAUV to navigate closer to the ship to cover every point in the mesh, which in this case were 10 centimeters apart.

Robotic inspections often use rectangular surveys, going one strip at a time, much like mowing a lawn.  For large hulls, however, this approach can get time consuming. Hover and Englot took a different, more efficient approach, writing optimization algorithms that program the robot to sweep across the structures, accounting for their three-dimensional shapes. “Over a minute or two of computation,” says Englot, “we’re able to make tremendous improvements to the length of this path, and do so while keeping every single point in view.”

The team has field tested its algorithms, creating underwater models of the Curtiss, a 183-meter military support ship in San Diego, and the Seneca, an 82-meter cutter in Boston. The researchers are performing further tests this month in Boston harbor.

The following video tells more about the project.

• Small-Scale Surveillance/Emergency Robot Developed
• Robot Jellyfish Runs on Hydrogen, Oxygen in Water
• Algorithms Plot Optimal Autonomous Underwater Vehicle Routes
• University Develops, Licenses Autonomous Robotics Technology
• Sensitive Underwater Microphone Works at All Depths

*     *     *