Under different working conditions, the tribological characteristics of materials show a complicated and non-linear relation. As a result, it is crucial to advance tribology by prioritising a data-driven strategy to estimate service capability in order to expedite the material design and preparation. With this aim, the present work firstly deals with the implementation of novel deep learning technologies in predicting tribological characteristics of additively manufactured and casted 316 stainless steel against 100 cr6 alloy. The coefficient of friction and frictional forces data from ball-on-flat experiments were used to develop the different deep learning models i.e., CNN, CNN-LSTM, and ATTENTION based CNN. Then, the wear tracks of tested samples were analysed with the SEM analysis. According to the findings of the wear rate, the AM material wears with an average of 58% less intensity than the casted material. In addition, the performance of the CNN Attention model demonstrated higher levels of accuracy and lower loss metrics in comparison to the CNN and CNN-LSTM classifiers.