Water wizards

Water distribution systems are aging, and under increasing stress from both droughts and severe rainfall, driving demand for AI-based solutions. The Tucson Water Department, for instance, uses a machine learning solution to conduct continuous virtual condition assessment and predictions for more than 4,600 miles of distribution water main pipes, assigning a failure risk score to each segment of pipe. Globally, AI has the potential to aid in the design, operation, and maintenance of treatment plants and distribution networks. Additionally, it can help manage water demand and promote water justice. However, the rapid deployment of AI in this context may introduce risks such as system-wide compromises due to design flaws, malfunctions, and cyberattacks.

As AI-enabled approaches to water system management spread to cities globally, it will bolster their adaptive capacity to withstand both prolonged droughts and sudden, intense storms that can overwhelm water system infrastructure.