In the context of Industry 4.0, developing both critical thinking and technical competence among engineering students is crucial to meet the demands of intelligent manufacturing. Traditional instructional methods face challenges such as limited access to equipment, high costs, and low engagement. To address these, this research presents a Virtual Simulation Teaching System (VSTS) for Computer Numerical Control (CNC) technology, integrating realistic CNC simulations and digital twin replication for immersive learning. Learner interactions, task performance metrics, and assessment results were systematically collected in the 187 CNC learner interaction and performance dataset from Kaggle, capturing time-series sequences, code entry, operational errors, toolpath scores, hints, idle time, and self-assessment. The dataset supports predictive modeling and personalized feedback using a Synthetic Feeding Bird fine-tuned Residual Convolutional Long Short-Term Memory (SFB-Res-ConvLSTM) network, which models spatial-temporal learner behaviors, predicts performance, quantifies uncertainty, and provides adaptive guidance. Contextbased instructional design embeds critical thinking in interactive machining tasks. A 6-week pilot study involving 187 CNC students employed paired t-tests and Cohen’s d to evaluate system effectiveness. Results indicate significant improvements in cognitive engagement (+21.25, p < 0.001, d = 1.64), selfconfidence (+1.29, p < 0.001, d = 1.45), CNC concept comprehension (+21.17, p < 0.001, d = 1.80), programming accuracy (+16.21%, p < 0.001, d = 1.64), and task completion time (−222.38 s, p < 0.001, d = −3.24). The SFB-Res-ConvLSTM achieved 98.7% prediction accuracy, RMSE of 0.078, MAE of 0.064, CI coverage of 97.3%, and PIAW of 0.112. Findings confirm that the integration of virtual simulation, deep learning, the CNC-VSLIPD dataset, and statistical evaluation yields a scalable, intelligent, and Industry 4.0aligned CNC teaching system, advancing digital transformation and cognitive skill development in modern engineering education.