University of the West of England and WASE Limited KTP 24_25 R2

Wastewater treatment plants are about to get a brain that learns on the job. Researchers at the University of the West of England are building a machine-learning "soft sensor" — a software platform that predicts and controls the behaviour of a biological reactor without needing physical probes inside the tank. The reactor in question is WASE’s Electro-Methanogenesis Reactor (EMR), which uses microbes and a small electric current to break down organic waste. Currently, scaling up this technology is tricky because the internal chemistry shifts unpredictably. The new platform will monitor conditions in real time, learn from the data, and adjust the process automatically. If it works, the impact is straightforward: businesses could treat their own wastewater on-site, cutting the cost and carbon footprint of trucking waste to central plants. The reactor would simultaneously produce biogas for energy, clean water, and nutrient-rich fertiliser. For a food factory or a brewery, that means lower bills, less pollution, and a local source of power and fertiliser. The project is applied engineering — it aims to turn a promising lab-scale reactor into a reliable commercial product that quietly keeps industrial wastewater out of rivers and sewers.