AI-optimized pulsed electric field for low-carbon urban water disinfection.

Urban water systems are critical hubs for the flow of carbon and materials in urban metabolism. Traditional disinfection methods, such as chlorination, face a dual challenge. On one hand, they rely on energy-intensive chemical production and transportation, resulting in significant indirect carbon emissions. On the other hand, they release disinfection by products that harm ecosystems and public health, which directly contradicts the "One Health" framework. To address these issues, we have developed a new disinfection platform centered around an AI-optimized pulsed electric field system. Using wide-bandgap semiconductor technology, the platform achieves high-frequency nanosecond pulse generation with up to 97% energy efficiency. Building on this, we conducted a systematic study on the inactivation mechanisms of resilient Bacillus and developed a key quantitative model. This model served as the foundation for building a dynamic AI adaptive control system that responds in real time to changes in water quality. Compared with fixed-parameter operation, the AI-guided mode maintained a high sterilization rate (72.3-81.8% inactivation) while reducing energy consumption by 15-25% and cutting performance variation by 47% in simulation tests. As a chemical-free process, this method eliminates the carbon footprint associated with traditional disinfectants and prevents the release of toxic byproducts. By combining precise physical disinfection with adaptive information technology, this study presents a sustainable, low-carbon solution for urban water management. Thus, the work may represent a step toward optimizing the flow of carbon and materials in cities.