Autonomous Underwater Robot Identifies Dangerous Naval Mines
Ryan Eustice, associate professor of naval architecture and marine engineering at the University of Michigan, has worked with Bluefin Robotics on an autonomous underwater vehicle (AUV) that conducts optical visual mapping and inspection of a ship's hull. High-resolution sonar images can be viewed in real time over the robot's fiber optic tether. The AUV can identify limpet mines on a ship, saving human divers from doing the dangerous job and doing away with present-day mapping and inspection methods that are time-consuming and imprecise.
Transcript
00:00:00 so what we did as a need experiment kind of really showcasing the capability of uh using this advanced technology and the objective here was within like 10 minutes we wanted to be able to drop the robot in the water have it figure out where it is relative to the to the map that had it had built in May and then drive back to an interesting feature that we had identified in the hall from
00:00:22 our May experiments what we're trying to do is develop robotic systems that are able to auton iously map ship holes while they're in Port for the specific problem of limpid Mines this is something that traditionally you would do by using human divers but it can be a dangerous and complex environment for a human to being especially if we're talking about
00:00:44 limp and mines that might be placed on the hole and so the idea is how you can remove the human from being at risk by having a robotic system do this so by equipping the robot with these camera systems what we're able to do is essentially allow the robot to nav navate much more like you or I do as a human in the sense that it can use visual information um to recognize
00:01:06 places on the on the shiple that it seen before so what we did is we deployed the robot um in May of this year on the USS Saratoga which which is a decommission aircraft carrier using our algorithms were're able to basically build this High Fidelity kind of three-dimensional map that represents kind of the blow water portion of the ship hole and then several months later in August we
00:01:26 dropped the robot back in the water and immediately like oh hey I recognize I've been in this hole or I know where exactly where I am and was able to snap itself back into that map and then immediately drive back to uh an interesting feature on on the hall one of the kind of cool things about the technology we've been developing is that if you really wanted
00:01:48 to map a hole fast the way you would do it is not with a single robot but with multiple robots all at the same time so I think ultimately the future application of this for the Navy would be not just a single robot that is being used to map a ship but probably teams of robots all working collectively together to do this the real kind of growth area for this is commercial applications of
00:02:10 this how do we assess the quality of the hole is there any damage to the ship is there any bio growth on the ship there's a lot of growth area for that in terms of the commercial applications of this that will outgrow from the Navy specific needs of I was trying to do the uh kind of liit mind detection on the hle that