Before we built anything, we cleaned the river by hand. Most of what we pulled out wasn't plastic.

Before designing the interceptor, we wanted to know what we were actually trying to catch.

So we ran a manual cleanup on the river. Volunteers, gloves, bags, a stretch of riverbank. We sorted everything we collected. Counted the bags. Weighed them. Categorized.

The results surprised us.

85 to 89% of the trash we collected was non-recyclable. And the largest single category wasn't bottles. It wasn't bags. It wasn't styrofoam.

It was textiles. Old clothes, fabric, rags. Waterlogged, sediment-filled textiles, sitting on the bottom of a creek in Asunción.

That changes the engineering problem completely.

Plastic is buoyant. It floats on the surface. A floating barrier catches it. That's the model most famous river cleanup projects are built on.

Textiles don't behave that way. Soaked clothes absorb water, fill with sand, and sink. They settle on the riverbed. A floating barrier catches zero percent of them.

So our interceptor needs a submerged component. Not optional. Built in from day one.

There's a bigger story underneath this finding — about where all those clothes came from, why people are putting them in the river, what the consumption-and-disposal cycle in this neighborhood actually looks like. We don't have full answers yet. But "the river is full of clothes" is not the same problem as "the river is full of plastic," and the solution can't be the same either.

We started thinking we were solving a plastic problem. The data is telling us it's at least partly a fast-fashion problem.

That's a useful thing to know before you start welding.

Next post: the actual interceptor design — three stages, no motor, recycled materials where possible.