AI-Enhanced Odour Monitoring for Wastewater Treatment

Use Case:

Continuous real-time monitoring and detection of odorous chemical emissions in wastewater treatment plants.

Outcome:

• Real-time detection of odorous compounds with sub-ppb sensitivity;
• Dual-prototype deployment in operational field sites (Rome plant) covering sedimentation and sludge drying zones;
• Alerts for hazardous odour events delivered via an online dashboard.

Ecosystem Support:

• Collaboration with AI mentors under the EU StairwAI project;
• Participation of local industrial end-users and technical opinion leaders;
• Iterative field testing and prototype co-design with stakeholders.

AI Relevance:

Demonstrates AI-driven enhancement of environmental and public health monitoring. Contributes to the European digital transition and sustainability goals by offering a replicable, high-impact AI use case.

Summary:

Pollution Srl, a technology SME based in Italy, has developed an AI-powered monitoring system to detect harmful odorous emissions in wastewater treatment plants. Collaborating with AI mentors under the EU STAR2OC project, the company addressed a critical gap in the sector: the lack of reliable, real-time odour detection tools. Traditional approaches—such as dynamic olfactometry and IOMS (Intelligent Odour Monitoring Systems)—either lack online capabilities or provide inconsistent results. Pollution Srl overcame this with a novel micro-GC analyzer coupled with a robust AI inference engine. The system continuously monitors emissions, identifying not just the presence of odorous compounds, but also providing confidence intervals and source localization. The AI model is cloud-trained and updated remotely, allowing adaptive learning from real-world data. Field tests were conducted in Rome’s central wastewater facility, impacting over 5,000 residents and workers. Two distinct prototypes were deployed: one in an open-air sedimentation zone and another in a high-impact sludge drying area. Results showed accurate detection of compounds like benzene, toluene, and xylenes, even at sub-ppb levels, with improved sensitivity and fast cleaning cycles. This innovation empowers plant operators and local governments to take action based on reliable data. The system enables early alerts to mitigate public health risks and reduce complaints, aligning industrial practices with environmental and societal needs. Thanks to AI integration, the solution is scalable to other sectors such as landfills, agriculture, and petrochemical industries.