Algorithms improve capabilities of underwater search robots

Massachusetts Institute of Technology (MIT) engineers have improved autonomous robots that search for underwater mines attached to ship hulls.

According to a statement, the mechanical engineers have designed algorithms that vastly improve the robots’ navigation and detection capabilities.

The algorithms enable the hovering autonomous underwater vehicle (HAUV) to move around a ship’s hull and view complex structures such as propellers and shafts.

The goal is to achieve a resolution fine enough to detect a 10cm mine attached to the side of a ship. 

‘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,’ said project leader Franz Hover from MIT’s Department of Mechanical Engineering. ‘The ability to ensure that the bottom of the boat doesn’t have a mine attached to it is really critical to vessel security today.’

Fully viewing a massive structure such as a naval combat vessel — as well as all its small features, including bolts, struts and any small mines — is a tricky planning problem, according to Hover.

‘It’s not enough to just view it from a safe distance,’ said Hover. ‘The vehicle has to go in and fly through the propellers and the rudders, trying to sweep everything, usually with short-range sensors that have a limited field of view.’

The researchers initially programmed the robot to approach the ship’s hull from a safe 10m distance, moving in a square around the structure.

The vehicle’s sonar camera emits signals that reflect back as the robot makes its way around the ship. However, at this distance, the signals are only able to generate a grainy image with a resolution so low that a small mine is not visible.

Furthermore, Hover said the misty image doesn’t necessarily tell a robot where a ship’s structure begins and ends — crucial information for the robot to avoid colliding with a ship’s propellers.

To translate this ‘mist’ into a solid structure, the researchers adapted computer-graphics algorithms to their sonar data, generating a three-dimensional mesh model.

The researchers then programmed the robot to swim closer to the ship, navigating around the structure based on the mesh model. Hover explained that the idea is for the robot to cover every point in the mesh; in this case, each point is spaced 10cm apart and is therefore narrow enough to detect a small mine.

The team has tested its algorithms in the field, creating underwater models of two vessels: the Curtiss, a 183m military support ship in San Diego, and the Seneca, an 82m cutter in Boston. The group is performing tests this month in Boston Harbour.

‘The goal is to be competitive with divers in speed and efficiency, covering every square inch of a ship,’ said Brendan Englot, a graduate working on the project. ‘We think we’re close.’

The research is supported by the Office of Naval Research and the findings were published in the International Journal of Robotics Research.