Applying ROS Industrial in an automated welding factory

Bison code: L.28355

Context/current situation:

The current trend in building is the increased use of steel beam structures and also an increased complexity of these structures. This allows the prefabrication of the structures in a factory and efficient assembly at the building site, which increases the efficiency of the building process.

For high mix low volume/one piece flow production of steel beam structures, companies rely on welders. Manual welding forms a bottleneck in the production process, which is worsened due to the decrease of the availability of well trained and experienced welders. Together with the higher demand and increased complexity of steel structures, this forms a challenge for many manufacturers. Many manufacturers want start using automated welding systems. Unlike welding systems used in series production (e.g. car industry), these systems need to be highly flexible with minimal programming effort to deal with the one piece flow.


Problem statement:

A producer of steel machinery has developed an automated welding system for steel beams that consists of multiple welding robots, a manipulator robot, and beam rotation system. Due to the high variety of products, increased complexity of steel structures and the tolerances of the components, the producer is developing intelligent software that can face the challenges that are imposed by this. The research group Mechatronics is looking into possibilities of ROS Industrial as a software framework to deal with some of these challenges. ROS Industrial is a software framework focused on developing software for the control of robots (e.g. articulated robots, AGV’s, flying robots, service robots), individual robots as well as multiple robots working together. This framework consists of many packages, for example handling sensor data, controlling the robot motors, path planning, navigation, mapping the environment. One of the main questions the research group tries to answer with this project is how ROS Industrial compares to other software technologies being used in industry.

Project or research goal:

Develop a simulation in ROS of the steel beam welding system and compare it with what can be done with technology being used in the existing system. This simulation tackles problems like:

  • path planning with collision detection of multiple robots;
  • dealing with deviations of part dimensions due to production tolerances using (vision) sensor systems and AI;
  • conversion of CAM data to tasks for the welding system;

Project results:

  • Simulation of the system
  • Evaluation of technology provided by ROS compared to software currently being used in the welding system

Cluster: Industrial Automation