With a rapidly changing climate and the emergence of readily available satellite imagery, there exists both the need and opportunity to switch over from an interval-based to a needs-based approach for the maintenance of infrastructure assets – HACARUS’s innovation helps enable this shift. Current solutions for AI-based image analysis typically require large, well-defined data sets for creating AI models. In addition, these often fail when applied to real-world scenarios due to sensitivity over discrepancies between data used for training and observed data. HACARUS’s solution is powered by its proprietary sparse modelling-based AI engine, which is uniquely adept at overcoming these challenges – the core algorithms allow for the creation of highly accurate AI models from small data sets, which means it is also able to quickly add new objects of interest. This allows for an AI system that is able to adapt to changes over time, and that can provide an instant bird’s eye view of an asset’s current health. The human-centric design, including smart features such as heat maps and bounding boxes, provides operators with actionable insights for smarter maintenance operations.