The production of high-performance rigid and lightweight composite materials is a top priority in automotive, defence, and aerospace industries. Therefore, it is crucial to introduce technologies related to Industry 4.0 to innovate the industrial production process. In the recent era, the use of an Artificial intelligence (AI) technology has exponentially grown and obtained significance as a powerful tool for simulating and modelling complex physical systems. Specifically, the autoclaving process facilitates the curing of composite materials of high-performance aerospace, automotive, and ships to get the desired strength and rigidness of the final product. The composite materials are subjected to high pressure and temperature to get durable, lightweight, and rigid products. Therefore, it is necessary to predict the actual temperature of an autoclave to obtain the desired strength and rigid products. In this work, we employed different machine learning (ML) approaches, namely, random forest, decision tree, gradient boosting, linear, multilayer perceptron, ridge, and balanced-ElasticNet regression for the prediction of the actual temperature of an autoclave. The elastic Net regression combines the penalties of both lasso and ridge regression and addresses the limitations of both. However, we introduced a balanced-ElasticNet by equalling both penalties to get the regularization and to handle the multicollinearity. The approach based on balanced-ElasticNet performs better compared to other ML approaches. Furthermore, we evaluated the performance using the historical data of 13 different batches and it obtained mean absolute error, root mean square error, R-2 squared error, and temperature relative error of 1.95, 5.71, 0.90, and 0.05, respectively. We also made a comparative analysis using different machine-learning approaches to check the reliability of approaches for accurate prediction of the actual temperature of an autoclave. However, the comparative analysis confirms the reliability of the balanced-ElasticNet-based approach for accurate prediction of an autoclave’s temperature. Furthermore, the proposed approach can assess, monitor, and improve the curing production processes of Dallara, which can lead to the production of the safest and most reliable lightweight and rigid products in the world.