Data-driven modeling to predict the load vs. displacement curves of targeted composite materials for industry 4.0 and smart manufacturing

•A machine learning-based approach for fiber-reinforced polypropylene composites design.•An artificial neural network to predict the load–displacement curves for targeted composite materials.•Reduce time and effort of the material designers for intelligent product design.•System identification through sparse identification of nonlinear dynamical systems.•An approach for smart manufacturing focusing on Industry 4.0. This work presents an approach for smart manufacturing focusing on Industry 4.0 to predict the load vs. displacement curve of targeted cotton fiber/Polypropylene (PP) composite materials while complying with the required intended properties. Experimental data for varying composite fiber percentage to characteristic load and earlier built artificial neural network (ANN) models are used as the feed. A newly built ANN model is being trained and tested on the TensorFlow backend using the Keras library in Python to predict the load vs. displacement curves for any in-between values of the experimental range (e.g., 0–50% cotton fiber filler content in PP) without doing any further experiment. Finally, a Python package for the sparse identification of nonlinear dynamical (PySINDy) systems is used to identify the exact data-driven ANN model through the system identification, which will facilitate the effective implementation of the control algorithms, smart internet of things (IoT), and high-tech automated system.