Machine Learning allows a fully automated compliance to local context & patterns, without requiring statistician or experts to analyse the historical records, or the cluster of patterns of other contexts. It adapts automatically to site perimeter evolution, new patterns, evolution of systems over-time and new contexts evolution, as well as aggregation. This also allows a network effect of automated learning from multiple patterns encountered across sites, for more precision and accuracy, avoiding false alerts.

Crystal is already deployed in a resilient and proven fashion, in heavy industrial contexts, managing large energy power in real-time.