From reactive to predictive. From assumptions to insight. HYD-ai helps you stay ahead of the flow.
HYD-ai is a Model-as-a-Service solution that leverages Hybrid AI to assess and quantify hydrate blockage risks in offshore oil production wells — in real time.
HYD-ai empowers flow assurance engineers and production teams to make proactive decisions, avoid unplanned shutdowns, and reduce the cost and impact of hydrate-related flow interruptions.
Developed in collaboration with Petrobras through the Conexões para Inovação program, HYD-ai was trained using a comprehensive dataset of real operational data from dozens of wells. The result is a robust, field-tested model designed to support production teams in high-risk flow assurance environments.
🔁 Hybrid AI: Physics + Data
Unlike traditional black-box models, HYD-ai combines real sensor data with simulated multiphase flow behavior, including hydrate formation, transport, and deposition. This hybrid approach enhances both accuracy and interpretability.
📊 Continuous Risk Assessment
HYD-ai continuously production conditions and evaluates the risk of hydrate blockage — offering operators early warnings monitorsand actionable recommendations.
🧠 Explainable & Traceable
With a built-in software platform, users can manage datasets, track model performance, and gain visibility into key drivers and root causes through intuitive explainability tools and what-if analysis.
⚙️ Field-Ready Integration
Designed to operate as a Model-as-a-Service, HYD-ai can be deployed rapidly and integrated into existing digital ecosystems or data platforms.