Determining the friction and wear behaviors of aero-engine key components under realistic conditions is important to improve their long-term reliability and service life. In this paper, the friction and wear behaviors of different bushing materials in the variable stator vane (VSV) system were investigated through the basic pin-on-disc test and actual shaft-bushing test. Different machine learning (ML) models were established based on the experimental information to predict the coefficient of friction (COF) and wear-rate. The results indicated that there is a significant temperature warning line for the wear amount of the polyimide material, while the high-temperature alloy material exhibited stable tribological performance under experimental load and temperature conditions. ML analysis indicated that the extreme gradient boosting (XGB) outperformed other ML algorithms in predicting the COF (R2 value = 0.956), while the kernel ridge regression (KRR) produced the best performance for predicting the wear-rate (R2 value = 0.997). The tribo-informatics research for bushings in the VSV system can accelerate the structural optimization and material selection and support the evaluation of new structures and materials.