Automation of the heavy metal industry2020-10-07
Digitalisation has proven to be an effective way to reduce the number of errors in the production stage. The new technology has mainly been used in, for example, the car manufacturing industry. In a new project, researchers at Karlstad University are looking at the really heavy metal industry where manual work is still common.
Forging processes in the heavy metal industry consist of many manual processes where an error may result in costly stops in production. Automating the various steps with the help of mechanical and materials engineering, secure data processing and the Internet of Things is the main focus of the research project “Smart Forge – sustainable production through AI controlled forging oven”.
- The digital transformation of various industries has already begun, says Andreas Kassler, professor of computer science at Karlstad University. However, in the metal industry they have not come very far yet. It is a heavy form of industry with many manual steps. In this project, we want to show the advantages of digitalisation and show the industry how they can benefit from it.
Bharat Forge Kilsta is a forging company with a great need to reduce the number of forged metal components that are discarded due to construction errors. The project “Smart Forge – sustainable production through AI controlled forging oven” is going to help them achieve that.
- An AI-based prototype system for automatic control of the forging line will be developed and tested. The hypothesis is that the system will be able to stabilise the process much faster after a restart than a manually controlled system and thereby reduce the number of discarded parts, Andreas Kassler explains.
The project will include the construction of a digital twin of the manufacturing process. It will be used to simulate how different heating methods affect the final product. Many of the people involved come from Karlstad University.
- We need expertise in materials science, forging and computer science to build the digital twin. Then it should be simulated and optimized using AI. Data coming from the production process must be verified to ensure that the AI-based decision-making processes are correct.
Are we witnessing a shift towards digitalisation even in the heavy industry?
- Yes, I think so. If something breaks in the production process, it is very expensive. With mechanical engineering, manufacturing processes can be improved, saving money as a result. That is a big change. Many companies see the economic advantages of digitalisation and are making big investments to increase their competitiveness.
Andreas Kassler sees the project as an excellent opportunity to fill the upcoming research centre at Karlstad University with activities.
- We want to run collaboration projects that focus on the digitalisation of production processes. Machine learning is a trend – if there is a process, it is possible to automate it.
The project will run for two years and has been granted five million SEK in funding from VINNOVA. Bharat Forge Kilsta, Viking AB and RISE are also part of the project.