Multiple robots can work on an assembly task in parallel and therefore make processes more flexible and efficient. The more robots interact, however, the more demanding their programming and coordination becomes. Indeed, particularly in the event of frequent changes in products or variants, adapting the systems becomes one of the key challenges.

© Fraunhofer IIS
Robotik Labor Fraunhofer IIS in Dresden
© Fraunhofer IIS, BLEND3 Frank Grätz

The research project DigiMon »Digital assembly with multirobot systems« explores how multirobot systems can be programmed, set up, and adapted to new tasks more easily. This work focuses on industry-ready multirobot technology, new teaching methods, and finding a no-code solution. The aim is to reduce the resources spent on engineering and facilitate the use of systems of this kind in industrial assembly processes, allowing companies to integrate new products or variants into existing production systems much more quickly.

A key first step: The technological building blocks are in place

With the project having been up and running for around 20 months, key findings from the conceptual phase are now available. A dual-arm robot platform with two collaborative robots has been set up and deployed, and a digital twin has been created for the virtual planning and testing of multirobot processes.

Another key aspect of development was a Robotics Toolbox that allows robot movements to be recorded, managed, and reused for new applications. An initial demonstrator has also been implemented for a selected assembly process.

So far, the project has demonstrated that these technological building blocks can be used together and provide a sound basis for further system development. The key task is now to refine this basis into an overall system that can be used to describe and teach complex robotic processes as intuitively as possible.

From feasibility to intuitive robot programming

The technological building blocks that have already been put in place provide a solid foundation. In the next phase of the project, the focus will now be on intuitive teaching methods and further system integration. This will be accompanied by an exploration of AI-based teaching methods with a view to more robust automation of even complex joining processes.

So far, the results show that successful use of multirobot systems relies not only on high-performance robots but also, above all, on suitable methods for programming and adaptation.