Robot Subs Compete


These Autonomous Underwater Vehicles are all competing in the 12th annual AUVSI competition. They have to complete an underwater obstacle course that involves some tight maneuvering, retrieval of a briefcase, dropping bombs, and firing torpedoes. We’ve seen several UAVs before, but we haven’t ever seen them weaponized and in action. Yeah, those weapons don’t look lethal, but isn’t that just a matter of ammunition?

9 thoughts on “Robot Subs Compete

  1. I actually went down there on the last day of competition (here in SD). It happened at a Marine underwater-acoustics facility, which was pretty nifty, but the real stars of the show were the bots.

    Cornell’s bot was an impressive beast, fit for underwater research to be certain — it was also freaking huge, in person, compared to most of the other bots. U of Maryland (iirc) had an impressive entry, and they were the last team to compete (being last year’s winners), but according to a member of the team they had a mis-detect of some variety, resulting in an incredibly off trajectory — when the bot corrected, it ran into the wall of the pool and just tried to keep pushing forward. Their mics couldn’t detect the beacon for the final goal (the “briefcase” that they had to recover), so it didn’t know how to course correct. It was sad, just watching it sit in one place.

    What I found the worst, though, was the “San Diego iBotics” (mostly UCSD students) team’s entry’s story — apparently they were sitting pretty, had a couple good partial test runs, and while they were tweaking settings and testing their system, they had a component failure (fancy compasses apparently !like heat, and they were running it for a long time out of water — aka, uncooled). That went belly-up, and their bot ended up going in circles. It was an impressive entry, anyway — nice carbon-fiber manta-ray design and everything. I was sad I didn’t get to see it compete.

    Oh well, maybe next year.

  2. This was such a blast to compete in, our bot is a first year bot this year so we only placed somewhere around the middle but we’re already looking forward to next year. I’m from NC State and I can attest to the difficulties of the competition. We started the competition trying to configure our compass and after a few days of work successfully had it up and running (it seemed we could do everything but go straight…) Prior to our final qualifying run everything was working smoothly (at least for the goals we were trying to complete) but after fixing bugs between our test run and final qualifying run we accidentally bug fixed a bug that was adding an offset to our yaw thrusters and stopped us from moving forward, at all. In the last minutes of our 15 minute qualifying run our main programmer managed to diagnose and fix the bug by adding a constant offset but also accidentally commented out a line of code that got us stuck looking downward for the plank to line us up to the next goal after getting under the gate.

    After a 10 month design cycle from concept to reality we’re happy with how far we’ve gotten and are looking forward to next year.

    Just a quick shout out to Virginia Tech, you guys were awesome looking forward to seeing you next year!

  3. @gert
    The ones that I have had experience with (PAP-104) use a fibre-optic umbilical, but those are remotely operated. I would assume that these are only tethered when they need external power or debugging.

  4. This is a fully autonomous competition. The competition is designed in such a way that the bots are forced for use markers to navigate from one mission to the next (basically orange boards that are placed in the water that run long ways in the direction of the next target) vision systems handle identification and response to the markers. SeaWolf employs a 3 camera system, one looking forward for seeing objects in front of us, one looking down for identifying the mission markers and “bombing run” mission markers, and one looking up to identify when we’ve successfully passed under the gate. We also use a depth transducer to avoid surfacing and a four hydrophone acoustics system for locating the pinger mission. During the practice runs you are allowed to tether for debugging purposes but during competition it’s a press start and pray kinda thing. These are Autonomous Underwater Vehicles (AUV) not Remote Operated Vehicles (ROV).

  5. @kgun some teams employed DVL’s but for the most part teams don’t have it in their budget for that kind of hardware. We’re working on an interesting system that will allow for velocity measurement sans any acoustics and any real cost. We plan to integrate that data and hopefully this will result in an accurate ability to “map” where we are in the water.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.