This research presents an AI-supported approach for fatigue assessment of SAE 4140 steel using non-destructive testing (NDT) methods, including infrared thermography, digital image correlation (DIC), and electrical resistance measurements. Real-time data is analysed using machine learning techniques to identify early-stage fatigue indicators and support the development of predictive fatigue models. The aim is to enable faster, more efficient testing and condition-based evaluation of components under dynamic loading conditions.