Turbomechanica
Prescriptive analytics for maximizing industrial equipment life and uptime
Turbomechanica® is Mechademy's cloud-based platform designed specifically for industrial machines and processes, particularly rotating equipment like compressors, turbines, and pumps. It integrates proprietary physics-based performance models with state-of-the-art machine learning and deep learning algorithms. The physics models compare actual equipment performance against expected baselines, enabling early detection of impending faults. Simultaneously, the AI/ML models analyze vast historical and real-time data to identify patterns, correlations, and potential failure modes.
Turbomechanica's unique orchestration strategy allows seamless data flow between physics and machine learning models, enabling hybrid physics+ML digital twins. This integration expands the scope of early fault detection, provides richer insights into fault causality, and allows the use of physics-generated synthetic sensors within ML models. The platform generates prescriptive alerts, actionable insights, and automatically visualizes diagnostic data.
Overall, Turbomechanica empowers plant personnel with domain-rich predictive and prescriptive analytics to maximize equipment life and uptime. It promises 2-10% increased uptime, over 15x return on investment, and 10-20% higher efficiency across industries like LNG, refining, upstream, midstream, offshore, petrochemicals, and power generation.
| Title | Description |
|---|---|
| Core Tech | AI & Advanced Analytics |
|
Physics models
|
Proprietary physics-based performance models that compare actual equipment performance to baseline performance for early fault detection
|
|
Machine learning
|
Uses deep learning and machine learning models trained on historical plant data or simulated data to identify patterns, correlations, and potential failure modes
|
|
Diagnostics
|
Sophisticated orchestration that integrates physics and machine learning models, enabling expanded fault detection scope, richer causal insights, hybrid digital twins, and prescriptive alerts
|
|
Features
|
Extensive asset library, OEM-agnostic performance models, drag-and-drop setup, real gas equations, domain-informed feature engineering, model building at scale, model drift detection, integrations with popular ML/DL frameworks, and end-to-end model lifecycle management
|
Relative Business Impact
Sign up or log in to your Explorer Plus or higher plan to get access to all the 0 deployment references.
Last Deployment Year