Simulation of Multivariable Monitoring and Operational Fault Prediction Using a Digital Twin for ROVs in Subsea Installation Operations

• Pedro S. Barreto CEFET-RJ
• Milena F. Pinto CEFET-RJ

Resumo

The increasing complexity of offshore operations and the growing demand for efficiency, safety, and sustainability have driven the development of intelligent digital solutions. This study proposes a digital twin model for remotely operated vehicles (ROVs) applied to subsea installation activities, integrating multiple interdependent operational variables such as motor temperature, electric current, depth, and accumulated operating time. Using a synthetic Big Data set, the system applies multivariable monitoring, anomaly detection, and machine learning algorithms to perform real-time fault prediction. The results demonstrate the model's effectiveness in anticipating failures, generating contextual recommendations, and reducing unplanned downtime, thereby improving the reliability of subsea systems. Although validated with synthetic data, the methodology is designed for straightforward adaptation to real-world sensor data and can be extended to other offshore assets such as subsea trees, launch towers, offshore cranes, and subsea production systems. This approach supports ESG (Environmental, Social, and Governance) objectives by promoting safer, more efficient, and sustainable offshore operations.

Palavras-chave: Remotely guided vehicles, Trees (botanical), Robot sensing systems, Digital twins, Safety, Reliability, Sustainable development, Monitoring, Anomaly detection, Synthetic data, Digital Twin, Remotely Operated Vehicle, Anomaly Detection, Predictive Maintenance, Subsea Assets