In the ever-evolving world of shipbuilding, managing risks effectively is no small feat. As vessels become more complex, so do the challenges in ensuring quality and sustainability. Enter Ahmet Fatih Yılmaz, a mechanical engineering professor at Karabuk University in Turkey, who’s tackling this issue head-on. His recent study, published in the *Engineering Science and Technology, an International Journal* (formerly known as *Hacettepe Journal of Engineering and Science*), introduces a novel approach to risk management in shipbuilding, blending traditional methods with cutting-edge technology.
Yılmaz’s research focuses on integrating Failure Modes and Effects Analysis (FMEA), a long-standing risk assessment method, with machine learning, specifically Random Forest (RF) regression. The goal? To create a more objective, scalable, and predictive framework for identifying and prioritizing defects in shipbuilding.
So, what does this mean for the maritime industry? Well, imagine having a tool that can predict potential failures and defects more accurately, saving time, money, and resources. That’s precisely what Yılmaz’s hybrid model aims to do. By analyzing 489 documented defects from the construction of an LNG-powered fishing vessel at Cemre Shipyard, the model was trained to predict failure modes and their risk priority numbers (RPNs) based on four key factors: cost, time, frequency, and stage.
The results are promising. The model achieved high predictive accuracy, with a coefficient of determination (R2) of 0.9738 and a mean absolute error (MAE) of 1.3470. It identified deformation, inappropriate production, and defective part usage as the most critical categories, with time loss and frequency emerging as the most significant features influencing RPNs.
“This hybrid approach bridges expert judgment with data-driven intelligence,” Yılmaz explains. “It offers a scalable, objective framework for real-time quality control.” The potential commercial impacts are substantial. By improving predictive defect management, shipyards can enhance their quality control processes, reduce costs, and increase efficiency. Moreover, the model’s robustness and capacity to estimate risk reduction potential for high-priority failure modes make it a valuable tool for continuous improvement.
The integration of FMEA with machine learning also opens up opportunities for broader industrial applications. Yılmaz envisions a future where this framework is integrated with enterprise systems, enabling automated risk monitoring and continuous improvement. “The results demonstrate that combining FMEA with ML can significantly advance predictive defect management in maritime manufacturing,” he states.
For maritime professionals, this research highlights the potential of leveraging data and advanced analytics to drive sustainable shipbuilding. As the industry continues to evolve, embracing such innovative approaches can pave the way for safer, more efficient, and environmentally friendly maritime practices.