Ai model for device control used in ocean for cleanup, animal rescue, recycling

Models
1

I’m thinking about finding some AI model that could be useful for visual detection of abandoned nets in the ocean and control of a device that untangles trapped seabirds, fishes, all kinds of animals stuck in abandoned fishing nets floating in seas and the oceans. Device would be possibly dragged under a boat, with camera, small scissors, able to auto detect loose floating net, check if animals are trapped- do some precision movement to free them with limited input from crew and ultimately compress the net and pass it to crew who would take those to recycle or repurpose.

So I’m curious to find best options maybe among these AIs now, IT HAS A LOT OF BENEFITS AND FUN TIME DOING THIS WHILE AND AFTER ITS developed into a Tech plus adventure pack- useful ocean exploration, CRUISING with MAGIC environmental benefit and fun and almost no extra work for the crew, that could really be great use of AI since not many divers want to do that work.

1 Like
2

To build what you described, you’d want:

  • Object detection model like YOLOv8 for analyzing video streams
  • A simple control loop (Python or Rust) to act on detections
  • Deployment on something like a Jetson Nano or Coral TPU for low-cost edge compute

If you want it to “learn” over time, you can add:

  • A feedback mechanism (e.g., log successes/failures and adapt thresholds or routines)
  • Optional cloud syncing if you need shared memory across devices

Keep it modular and don’t overthink the AI part yet. Start small, iterate fast.

1 Like