Machine learning framework for wastewater circular economy - Towards smarter nutrient recoveries

As the world's supply chains become disrupted through geopolitical instability and the race towards a net-zero future, policies have been implemented to improve the security of certain minerals and raw materials critical to a country's survival and sustainability goals. Circular economies (CE) are sought to be an ecosystem that will reduce virgin material consumption rates, lower carbon emissions, and decelerate the rate of landfilling. However, cost-effective and commercially attractive substitutes to conventional products are needed for this to be realised. Machine learning (ML) and the explosion of interest in artificial intelligence (AI) have led to growing interests in predictive and generative applications for sustainability. Phosphorous and, nutrients overall, operate on finite reserves essential for food supply chains; while such nutrients are largely present in municipal wastewater streams. Wastewater treatment plants (WWTPs) must then face a transformational force to become nutrient recovery centres, rather than follow a linear treat-for-disposal model. In this framework paper, ML is positioned as an enabler for scaled, cost-effective and safer recovery of nutrients and other valuable products - tying in economic, societal, technical and commercial factors through open data connectivity. Moreover, the paper issues a policy guide for institutions wishing to advance food, energy and water security through machine learning, circular economy wastewater treatment plants (ML CE WWTP).