Metal additive manufacturing is applied to diverse fields including aerospace, automotive and medical areas. It often benefits from sensor embedding for monitoring, among which fiber optical sensor embedding holds promise (Fig. 9). In particular, fiber bragg grating (FBG) sensors provide can detect both temperature and strain. This project aims to predict these information from FBG sensor data using machine learning (ML) models. It generates FBG sensor data via simulation. Among various ML models, nonlinear models such as Gaussian Process or MLP performs better than other models under noisy condition. Seong-Cheol explains that this result might be related to the fact that in realistic operating conditions, the FBG sensor data is expected to exhibit nonlinear behavior.