e-LAS+ – Multimodal quality assurance for production of electricity storage in safety-critical systems

roboter fertigen Batteriezellen
© adobestock.com/phonlamaiphoto

Description and project goals 


Initial situation and idea

Laser welding is a demanding processing method and has become an indispensable part of modern industrial production, for example in battery cell manufacturing. The challenge in laser welding is to generate an optimal heat application by the laser so that a reliable joint is produced by the weld seam to avoid damage to the cell. Due to the complexity, the determination of the optimal process parameters is currently carried out via a very elaborate, iterative process, with numerous destructive tests. In order to meet the high requirements in lightweight construction, to achieve good welding results and at the same time to achieve high productivity, many different influences in the process must be taken into account. In this context, process monitoring plays a major role.


About the project

The e-LAS+ project focuses on the development of a multimodal inline inspection process for laser welding. Furthermore, with the help of a predictive process adaptation as well as an AI-based platform, the economic operation of laser welding equipment for small batch sizes is to be enabled. The systematic recording of the experience values of the machine operators will be used to be able to improve the complete value chain.


Innovation in automated monitoring

The planned multimodal inline inspection process is based on machine learning (ML) methods and ensures the quality of the weld seam for small batch sizes.  Occurring defect patterns and process regulation options will be made available to the machine operators as a digital assistance system in order to contribute to the future increase of product quality of electricity storage devices and to the efficiency of production. All project results will be demonstrated in a pilot plant. Furthermore, the developed inline testing method with assistance system will have the possibility to be used in the field of large-scale production.

Responsibilities of Fraunhofer IDMT

  • Development of a multimodal in-line measurement method using machine learning techniques (focus on airborne sound analysis)
  • Development of multivariate machine learning algorithms for multimodal quality assurance in the production of electricity storage systems


  • NeuroControls GmbH
  • SSI Software Service GmbH
  • SITEC Industrietechnologie GmbH
  • VRI GmbH


Federal Ministry of Economics and Climate Protection BMWK


May 2023 – April 2026

Industrial Sound Analysis – Research and Practice

"Research in Priority Funding Battery Cell Manufacturing" in the 7th Energy Research Program of the German Federal Government.