Sensing and Accounting of Aquatic Floatables · IIDA Labs
Waste reaching the sea is rarely measured. SAAF uses drone imagery, flown across tides and seasons, to show how much waste there is, what it is and where it gathers.
Request a demoMost interventions are planned without baseline data. SAAF builds a needs assessment for places where the waste entering the system goes uncaptured: its weight, its categories, its hotspots and its seasonal trends. Faster, more repeatable and wider in reach than a manual audit, and flown again to show whether the work upstream is paying off.
Where it works
Waste washing ashore is unmeasured.
Captures the weight and category of waste the sea deposits on the shore, across tides and seasons, giving a clear picture of scale and pattern.
The problem in fullWaste moves in both directions.
Captures floating waste on the banks by quantity and category at different tide times, revealing flow patterns and accumulation hotspots.
The problem in fullTheir contribution is invisible.
Measures each city's share of the overall waste chain, making upstream contributions visible and supporting shared responsibility along the river.
The problem in fullThe common thread. In all three, the gap is missing data. SAAF turns an invisible problem into a measurable one, which is the first step toward targeted action.
00The problem
So clean-ups are planned on impressions: a photo, a volunteer's estimate, a short stretch counted by hand and multiplied up. Without a baseline, no one can say whether the problem is growing or shrinking.
river mouth · floating raft
01Fly the survey
Repeat flights show how waste moves and where it accumulates. No special sensor: a consumer 4K camera at 10 to 40 metres is enough, because altitude and focal length give the ground scale.
10–40 m · 3840×2160
02Detect and classify
The model picks out each item and names its category. Six classes today: glass, plastic bottles, cans, takeaway cups, retort pouches, food containers. About a second per 4K frame.
6 classes · ~1 s per frame
03Measure and report
Item sizes pass through the ground sample distance, then sourced reference weights, to give an estimated weight per category. Mapped over tides and seasons, that becomes a needs assessment for the site.
count · footprint · mass
What the model sees
A real survey frame. 3840×2160, shot at altitude. Twelve pieces of litter, none of them obvious. Boxes are deliberately tight: a loose box inflates the area, and area is what the mass estimate is built on.
Why it is hard
This rectangle is a 3840×2160 frame scaled to your screen. The small box inside it is a 60-pixel object at the same scale. That is what the model has to find, name and measure.
Who it is for
Plan clean-ups where the waste actually gathers, and track progress in kilograms from one survey to the next.
Learn more
Target collection at the hotspots, by tide and by season, and show what changed once crews have been.
Learn more
Study leakage patterns with repeatable data across tides, seasons and sites. Same method every time, every frame retained.
Learn more
See verified impact: the same stretch measured before and after the work you funded.
Learn more
Send us a survey. We return the annotated frames and the totals before you commit to anything.
Request a demoQuestions
Where there is a limit, it is stated. Mass is an estimate from reference weights, not a weighing, and the site says so everywhere the number appears.
In three settings. Coastal cities, where it measures the waste the sea throws back onto the shore. Creeks and rivers that flow straight into the ocean, where high tides push waste in from the sea and low tides carry land-based waste out, so it measures the floating waste caught on the banks. And inland cities along major tributaries, where it measures each city's contribution to the waste that eventually reaches the sea.
Because waste moves. A single flight is a snapshot; flights at high and low tide, repeated across the year, show where waste comes from, where it accumulates and whether it is growing or shrinking. Over time that shows whether upstream interventions are reducing what reaches the sea.
Both are mAP, mean average precision, the standard score for object detectors: it combines how much of the litter was found with how few false alarms were raised, averaged over all six classes, on imagery the model had never seen. 100% is perfect. 99.2% is the everyday score, where a box counts if it covers at least half the item. 68.2% is the strict score, where the box has to fit almost exactly. The strict one is what the size and mass estimates depend on.
Each detection's box is converted to real dimensions using the ground sample distance, derived from sensor width, focal length and flight altitude. Those dimensions are matched against published reference weights for each class of item. The result is an estimate, not a weighing, and it is reported as one.
Six classes today: glass, plastic bottles, tin and aluminium cans, takeaway cups, retort pouches and food containers. New classes can be added.
A consumer drone with a 4K camera, and the altitude and camera model used for the flight. Nothing else. Imagery is uploaded and processed after the flight.
One survey. Annotated frames and totals back. Then decide.
Request a demo