Collaborative welding – humans and robots working in unison

TALENTED research project

Motivation and problem statement

MIT Moderne IndustrieTechnik GmbH develops custom system fittings, mounting and welding assemblies as well as special fittings. The introduction of automation solutions has not been appealing for the company so far due to low quantities, as these often require high investments. What is missing are efficient solutions that can also be used economically for batch sizes of 1. Fraunhofer IEM and MIT are demonstrating how Industry 4.0 in the form of flexible automation measures and collaborative robots (cobots) can also make its way into medium-sized companies in the TALENTED joint research project. 

Mann vor einer Maschine mit Greifarm.
© Fraunhofer IEM / Wolfram Schroll
Collaborative robotics for flexible welding

Project objectives and approach

As part of TALENTED, the project partners are developing a collaborative solution for what was previously an entirely manual welding process. The aim is to make the production steps, which are often performed manually in special machine construction, more efficient through intelligent automation measures. To accomplish this, sensor solutions are integrated into the tool head of a welding robot for MIT Moderne IndustrieTechnik. With the use of intelligent information processing, the image data is automatically evaluated and used to track the tool during the welding process. Using an control panel, employees only have to select the surfaces or edges to be processed on the 3D model of the workpiece and configure the welding process.

This does away with the time needed for programming each workpiece individually. The sensors record the contours of the workpiece, the system recognizes deviations from the design model and on its own compensates for them during processing.The process is partially automated, with humans and robots working together as a team. The collaborative robot arm is set up flexibly by the employee with no programming skills required. It frees him from having to do heavy physical work, but always leaves the employee in control of the welding process. This allows for a significant increase in the efficiency of the operation and a reduction in the susceptibility to errors and physically demanding work.

Project profile

Project title

TALENTED - Automatic and flexible self-monitored welding system for small batch sizes


March 2019 to February 2022


Canada-Germany 2+2 Programm - BMBF (DLR)

Project volume

approx. €2.1 million

Funding volume

approx. €1.3 million

cooperation partners
  • MIT Moderne Industrietechnik GmbH & Co. KG
  • Novarc Technologies Inc.
  • University of British Columbia UBC (Department of Electrical and Computer Engineering in the Faculty of Applied Science)
Project leader

Dr.-Ing. Christian Henke

The objective is to develop a robot-based collaborative welding system that automatically locates and precisely tracks seams to be welded. The welding system is designed for products with low and very low quantities, where there are also deviations between the real component and the associated CAD data, for example due to material fluctuations or welding distortions during the welding process. The aim is to minimize the effort required for adapting (setting up, calibrating) the system to the different workpieces, while completely eliminating the need for system- or robot-specific programming. In order to increase the quality of the resulting weld seams in addition to saving time during the adaptation of the welding system, self-learning approaches and self-optimization are also integrated.

Is this topic also of interest to you? Then feel free to contact us!

Christian Henke

Contact Press / Media

Dr.-Ing. Christian Henke

Head of Department Scientific Automation

Fraunhofer Institute for Mechatronic Systems Design IEM
Zukunftsmeile 1
33102 Paderborn

Phone +49 5251 5465-126