Roadheaders play a crucial role in the excavation processes of tunnels and mines, offering efficient and precise cutting capabilities. The performance prediction of a roadheader is essential for optimizing operations and ensuring project success. By understanding the various factors that influence performance, implementing predictive models, and continuously improving machine design and operational strategies, the potential of roadheaders can be maximized. This article delves into the intricacies of performance prediction for roadheaders, exploring methods, case studies, challenges, and future directions in this critical aspect of tunneling and mining operations. The primary objective of this study is to develop models that can predict the Instantaneous Cutting Rate (ICR), which is defined as the production rate during the actual cutting period (measured in tons or cubic meters per cutting hour), based on the properties of the rock formations being excavated as well as machine parameters. In this research, the Instantaneous Cutting Rate of roadheaders at the Tabas coal mine was analyzed by examining the characteristics of both the rock and the machinery involved. Additionally, this study employed Firefly Algorithm (FA), Bat Algorithm (BA) and Support Vector Machine (SVM), which were assessed using coefficient of determination (R²), root mean square error (RMSE), mean squared error (MSE) and mean absolute error (MAE).The obtained results for Firefly Algorithm (FA) are found to be as R2 = 0.9104, RMSE = 0.0658, MSE= 0.0043 and MAE= 0.0039, for Bat Algorithm (BA) are found to be as R2 = 0.9421, RMSE = 0.0528, MSE= 0.0027 and MAE= 0.0024, and for Support Vector Machine (SVM) are found to be as R2 = 0.8795, RMSE = 0.0762, MSE= 0.0058 and MAE= 0.0052, respectively. It can be concluded that while predictive models produce satisfactory results, the Bat Algorithm (BA) demonstrates a higher level of precision and realism in its outcomes.