How is Remaining Useful Life (RUL) estimated in data-driven Prognostics and Health Management (PHM)?

• A. By guessing based on experience.
• B. By training prognostic models on sensor data to predict time-to-failure or health indicators; common methods include machine learning and statistical techniques.
• C. By measuring only at the end of life.
• D. By summing the rates of component failures.

Answer: (B)