Methods for creating digital twins of car manufacturing plants

Team Members: Florian NoichlAndré Borrmann


Runnung period: 11/2019 - 10/2022

This project's aim is the automated generation of geometric and semantic models of car manufacturing plants, respresenting the actual state of the facility.

Project goals:

  • Preparation of a comprehensive concept for digital twins for car manufacturing plants
  • Development of methods for automating the evaluation of point clouds for creating semantically rich geometric models of car manufacturing plants
  • Development of robust CNN-based methods for the identification of semantic-geometric objects
  • Pilot run on an industry-relevant example

Our focus lies on structural and MEP components, with input point clouds mostly acquired with 3D laser scanners.

The project is conducted in cooperation with Audi AG, Ingolstadt.