The desal plant is learning its membranes
the desal plant is learning its membranes: ai now predicts reverse-osmosis fouling seven to fourteen days out, tracks energy-recovery devices against a 96 percent benchmark, and catches pump cavitation in the vibration spectrum. shenzhen’s leakage control, an 18-million-resident network, won the first ai project of the year at the global water awards.
the water is becoming a model with a maintenance schedule.
Context
Several papers cover machine learning for reverse osmosis desalination: RO-TRACK (2023) on predictive analytics for seawater plants where membrane replacement is a major cost, a 2025 ACS multi-model ensemble predicting RO performance and membrane fouling, a Feb 2026 comparison of regression and neural-network models, and a temporal fusion transformer for differential pressure.
Each treats fouling as the central operating problem.
Machine learning predicting reverse osmosis fouling matches the literature read. Prediction seven to fourteen days out, energy-recovery devices tracked against a 96 percent benchmark, pump cavitation caught in the vibration spectrum, and Shenzhen's leakage control winning at the global water awards were not seen in the sources read, so unsupported here, not refuted.
Related work
- Exa: RO-TRACK, data driven predictive analytics for seawater reverse osmosis plants ↗Plant analytics.
- ACS Environmental Science and Technology: Multi-model ensemble for reverse osmosis performance prediction ↗Fouling prediction.
- ScienceDirect: Temporal fusion transformer for predicting differential pressure in reverse osmosis ↗Pressure forecasting.
Watch next
- Find the global water awards entry for Shenzhen's leakage control.
Sources
- Exa, RO-TRACK, Oct 2023exa.ai
- ACS, Jul 30, 2025pubs.acs.org
- Exa, RO fouling with MLR and ANN, Feb 28, 2026exa.ai
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 2 October 2026 at 16:16 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →