Pipeline Risk

Pipeline Risk is Rezatec's likelihood-of-failure (LoF) modelling product for water utilities that need to decide where to send leak detection crews and which mains to replace first, rather than ranking pipes by age and material alone. For each utility, Rezatec trains a machine-learning model on the network's own break history and pipe attributes, combined with more than 100 environmental and geospatial datasets. The satellite inputs Rezatec names are C-band radar ground-deformation monitoring and multispectral vegetation indices (NDVI), used alongside soil, slope and weather data. The model scores pipe segments of roughly 100 m for Likelihood of Failure, Consequence of Failure and overall Criticality, and shows a Certainty Index and the main risk drivers (such as material, soil, ground motion or vegetation change) for each segment. Results are delivered in a secure web-based geospatial platform with interactive maps, filtering, messaging and alerts, and can be brought into a utility's existing GIS and mapping systems. Rezatec states that the model is validated for every customer, predicts over 70% of failures within the top 30% highest-risk segments, and helps leak detection teams find up to 5x more leaks per mile surveyed. Published customer results include WaterOne (Kansas), whose validation project found the model 78% accurate, and the City of Olathe, which used the outputs to raise its annual pipe replacement budget from $1.5 million to $6 million. Rezatec offers a free pilot before a full-scale programme.