Discuss a project
Menu

CASE STUDIES

Shellfish plot monitoring, round after round

Mussel and oyster growers judge a plot by dredging a sample or sending a diver, and the result lives in someone's head. This demo shows the alternative: the same plot inspected on the same lines every round, scored automatically, and trended against the water data, so the grower and the water authority both read the same page.

One season in 68 seconds. Simulation with real inspection footage and a drawn AI overlay; figures are illustrative.Marine RoboTech

What the workflow does

  1. 1. Map the plots without a crewed trip.

    A small uncrewed survey boat runs planned lines from the quay and returns a map of beds, anchors and lost gear. After a storm, the check takes a day.

  2. 2. Score the plot instead of sampling it.

    An inspection ROV films the ropes and the seabed on fixed lines. The footage is scored for coverage, size classes, seed density, predators, fouling and mortality.

  3. 3. Trend it round by round.

    Each round adds a point. Scores sit next to water quality, tide and weather, and a warning fires before the water turns.

  4. 4. One page for the season.

    The grower sees a score card per plot; the authority sees evidence without disturbance; one click shares the season report.

What is real and what is simulated

  • The survey boat, inspection ROV, cameras and sonar are standard equipment, available today (see Products).
  • The AI scoring shown in the video is a concept overlay drawn on real inspection footage, not model output. Scoring models exist in research for fish cages and mussel ropes; a regional dataset is built in the first season of a pilot.
  • The dashboard figures are illustrative.

How a pilot starts

One season, one or two plots, inspection and survey rounds on fixed lines, footage labelled with a knowledge partner and checked against the grower's own sampling. Season two automates the scoring; season three adds forecasting.

Discuss a shellfish pilot