Integrated system-level framework for inspection and cost-effective maintenance of jacket type offshore platforms using dynamic Bayesian networks

• System-level probabilistic framework integrating fracture mechanics and DBNs for offshore jacket inspection and maintenance. • Hybrid inspection strategies improve safety and cost efficiency. • Modular DBN approach allows scalable application to large offshore structures. This study presents a system-level framework for inspection and maintenance planning of offshore jacket structures subjected to fatigue deterioration. The framework integrates fracture mechanics–based crack growth models with Dynamic Bayesian Networks (DBNs) to probabilistically update structural states as inspection information becomes available, and it employs influence diagrams with heuristic optimization to identify cost-efficient inspection and repair strategies. The approach captures both component- and system-level deterioration, accounts for inspection outcomes in real time, and evaluates life-cycle costs by jointly considering inspection, repair, and failure consequences. Results highlight that purely periodic inspections may reduce cost but risk compromising reliability, whereas hybrid strategies that combine periodic scheduling with reliability-based triggers provide a more balanced and robust outcome. The framework is computationally efficient, scalable to large jackets with many joints, and adaptable to diverse deterioration scenarios. In practice, it provides a decision-support tool for asset managers and regulators to design inspection programs that balance safety, reliability, and economic efficiency, ensuring the long-term fitness for purpose of offshore infrastructure.