The articulated Heavy Earth Moving Machinery predominantly uses hydrostatic steering, because of its reliability and redundancy. In earlier studies, an energy efficient Electro-Hydrostatic Steering System was proposed, which works on the Steer-by-Wire principle and complies with the safety standards. In the proposed steering system, the redundancy of the steering is achieved by an electronically controlled proportional valve circuit that activates when a fault is detected in the steering operation. The detection of fault and activation of secondary steering is safety critical in the operation, failure of which may lead to a hazardous outcome. In this paper, an intelligent approach is taken to identify the fault in steering using pressure signals. Different algorithms based on machine learning and deep learning, namely bagged decision tree ensemble, multi-layer perceptron, and Gaussian kernel-based Naive Bayes classifiers are selected for this work. A real-time interactive co-simulation environment integrating the proposed electro-hydrostatic steering system is used for the study. Two fault scenarios corresponding to the major hazardous outcomes related to steering are carefully simulated to capture the fault conditions in steering, ensuring that the classifiers are trained on a diverse and representative dataset. Finally, an ensemble of all the trained classifiers is created and integrated into co-simulation model to detect the faults in real-time simulation, using probabilistic approach. The study demonstrates an effective use of artificial intelligence (AI) to ensure safety through redundancy in the Steer-by-Wire for heavy earth-moving machinery.