Estimating where human waste ends up in cities that have never been surveyed
- Good it would do
- 4 / 5
- Can AI help
- 3 / 5
- How neglected
- 3 / 5
- Can we check it
- 3 / 5
Why it's stuck
A 2020 analysis of 39 cities with 72 million people found 58 percent of human waste is not safely managed, but each city's estimate takes months of local fieldwork and only around 90 such reports existed. Most of the world's fast-growing small cities have no estimate at all, so money for sanitation is targeted blind.
What we'd try
Train models that predict each step of a city's waste flow from public data that covers every city: household surveys, satellite building density, sewer and treatment plant locations, and groundwater depth. Then estimate every city in low-income countries with uncertainty, checked by holding back existing city studies.
What could stop us
Only about 90 to 160 field-based city waste-flow studies exist, with inconsistent definitions and assumed treatment figures, so labels may be too few and noisy to train or test a model.
Source: https://frontiersin.org/journals/environmental-science/articles/10.3389/fenvs.2020.00001/full
Do you know the answer?
One reply from someone in the field can settle what our AI can't.