A Swedish research initiative is exploring how recycled wastewater can be used safely in food production. Monitoring water quality in these environments is notoriously hard: raw biological data is noisy, variable and expensive to analyse in a lab.

‍

We are building the project's data pipeline across hardware, software and AI. That includes a custom gas sensor array, an "electronic nose", and a machine learning engine that infers bacterial presence from gas signatures, turning noisy sensor readings into clear signals for decision-making.

Value for the client:

• Early detection of bacterial changes in the water
• AI-driven monitoring in place of slow, expensive lab cultures
• Noisy sensor data turned into clear visual decision tools
• Models that improve as more data is collected