Flexible Industrial Robots for a Resilient Industry
Project GANResilRob for the further development of production robots with generative AI successfully completed
Research Focus: Applied Artificial Intelligence
Industrial robots are precise, consistent, robust, and powerful, but not flexible enough to adapt to changing conditions in production. This has become very clear once again in recent years, with multiple crises. The project GANResilRob, funded by the German Federal Ministry for Economic Affairs and Energy, which has now been successfully completed after three years, addresses this challenge. Robots in industrial companies need to be both resilient and flexible in uncertain situations. The key to overcoming this shortcoming lies in generative Artificial Intelligence (AI).
The trend away from warehousing jeopardizes the seamless continuation of production processes and thus the end-to-end material supply chain due to delivery bottlenecks. Even if alternative suppliers are available, highly specialized industrial equipment must be adaptable quickly and, if possible, without additional experts. In most cases, the basic process does not change, but tools and workflows may need to be adjusted.
To meet this challenge, the project participants at the FZI Research Center for Information Technology developed an AI-based processing pipeline that allows robots to be quickly adapted and continue to be used even when processes change. In this way, they contribute to a resilient industry in terms of adaptive production processes.
Decisive innovation advance: Flexible robots through generative AI such as LLMs & VLMs
The FZI solution for the processing chain can automatically generate an efficient assembly sequence from an existing CAD (computer-aided design) model. The appropriate robot capabilities are selected for the necessary steps so that a robot can execute them directly. Camera-based recognition captures the actual assembly situation and combines this information with the assembly plan. A downstream AI module then creates a ready-to-use robot program – fully automated and flexibly adaptable to different products or components.
The decisive innovation advance: The intelligent use of large language models (LLMs) and vision-language models (VLMs) for the system enables rapid, flexible adaptation to changing tasks and conditions. The modified process does not require time-consuming manual programming. This achieves a new level of cross-process autonomy and industrial resilience, delivering real added value in dynamic industrial environments.
Use cases: Vacuum chamber assembly and disassembly of batteries for recycling purposes
The system equipped with generative AI was deployed and evaluated in two use cases. Both an assembly and a disassembly scenario were tested.
Use case 1: Flexible vacuum chamber assembly
The basis was provided by the experience of Pfeiffer Vacuum + Fab Solutions GmbH, an international mechanical engineering group specializing in vacuum pump systems and components. The FZI solution made it possible to derive handling steps for the assembly of vacuum chambers and to implement automation with two robot arms.
Vacuum chambers are often one-off and custom-made products. The flexible selection of assembly capabilities based on identified work steps enables a flexible assembly process that can be quickly adapted to modified derivatives.
Use case 2: Robot-assisted, automated battery dismantling as a contribution to the circular economy and safety
The FZI scientists deliberately chose the use case of automated dismantling of old electrical appliances, with a successful five-day real-world test at Electrocycling GmbH in Goslar. A system was set up and tested on site to dismantle heat cost allocators for recycling and to remove the batteries they contain. The FZI team was supported by regional partner REWIMET e.V.
Through technology transfer, the FZI researchers support economically viable recycling processes. In addition to recovering secondary raw materials, robot dismantling ensures the safe disposal of hazardous substances from lithium-ion batteries and minimizes the fire risk during further recycling. Automation makes this previously manual and dangerous step scalable, efficient, and safe.
Cooperation desired: exchange and communication on the topic
Do you work on or with flexible manufacturing systems, AI in production, or robot-assisted recycling? Would you like to exchange ideas or collaborate with us? Then get in touch with our FZI contact, David Timmermann.
About the FZI
The FZI Research Center for Information Technology, with headquarters in Karlsruhe and a branch office in Berlin, is a non-profit institution for information technology application research and technology transfer. It delivers the latest scientific findings in information technology to companies and public institutions and qualifies individuals for academic and business careers or the leap into self-employment. Supervised by professors from various faculties, the research groups at the FZI develop interdisciplinary concepts, software, hardware and system solutions for their clients and implement the solutions found as prototypes. The FZI House of Living Labs provides a unique research environment for application research. The FZI is an innovation partner of the Karlsruhe Institute of Technology (KIT) and strategic partner of the German Informatics Society (GI).
