Safeswim: Integrating Real-Time Monitoring and Modeling for Near-Term Predictions of Water Quality

Originally implemented in 1995 to communicate bathing water quality to the public, Auckland Council’s water quality management program provided an assessment based on retrospective water quality sampling. Safeswim (www.safeswim.org.nz) was deployed in 2017 to overcome limitations related to sampling data availability and to maximize the use of water monitoring data by integrating such data with real-time modeling, artificial intelligence, big-data analytics, visualization, and information dissemination. Safeswim is shaped around a single, central data platform Moata that gives Safeswim its capacity for managing and analyzing large, diverse datasets while running multiple models to provide Aucklanders with information updated in 15-min intervals. The platform has facilitated Safeswim’s greatly increased reliability and prediction performance over five years on from its first implementation, to close to 90% overall accuracy, and as a result has grown into a reliable and efficient public decision support system for the Auckland Council. The success of Safeswim in Auckland has led other communities around the world to consider and are inspired to implement similar programs.