Numerical data driven operation support for manufacturing of automotive body components

Post by: Tobias Larsson
3rd August 2025
No Comment

Abstract

With the increased focus on smart manufacturing and Industry 4.0, the use of simulations for the creation of cyber-physical manufacturing systems is increasing. The sheet metal forming manufacturing process, commonly used for production of automotive body components, is one of the processes that currently benefits from the use of simulations without exploiting them in a cyber-physical system setup. This study set out to initially identify the key controllable and uncontrollable parameters of the sheet metal forming manufacturing process for the design of an intelligent quality controller. Subsequently, the study investigates the possibility of using data points from a stochastic numerical analysis as training data for an Artificial Neural Network. The stochastic numerical model used is based on the existing Finite Element simulation standard at Volvo Cars to allow for a seamless integration of the methodology into the standard workflow of CAE departments. Lastly, the study will present a validation of the trained Artificial Neural Network using the Volvo XC90 inner front door component as an industrial demonstrator.

Keywords

Artificial neural network, Deep drawing, Process control, Virtual shadow, Industry 4.0

Reference

Barlo, A., Aeddula, O., Sigvant, M. et al. Numerical data driven operation support for manufacturing of automotive body components. J Intell Manuf (2025). https://doi.org/10.1007 /s10845-025-02664-8

Download

https://doi.org/10.1007/s10845-025-02664-8

More info

Sharing

Categories: Publications, Research

Tagged: Artificial neural network, Deep drawing, eureka, johan pilthammar, mats sigvant, md3s, mdd, Omsri Aeddula, Process Control, SDD, shafiqul md islam, smart, Virtual shadow