SeConRob is an innovative project focused on developing self-configuring methods for robotic workflows and processes in manufacturing. This initiative addresses the challenge where each step in the manufacturing process depends on the outcome of the previous one. Such processes are currently inefficient in terms of productivity, energy, and resource utilisation due to limitations in automation, for technical and economic reasons. This inefficiency is particularly obvious when it comes to inspection and subsequent re-work stages, where the re-work is dependent on the results of the prior inspection for each individual part. SeConRob's goal is to introduce technologies that facilitate the automation of these stages. The project envisions creating self-configuring robotic processes, tailored for each part, based on AI-driven analysis of the inspection data. This analysis will be used to automatically generate robotic programming and process parameters for the re-work stage. Additionally, physical process models will serve as the foundation for initial planning, complemented by a long-term reinforcement learning based feedback loop to enhance the process and incorporate aspects not covered in the initial model.
The project will test its advancements through two use cases involving multi-stage manufacturing processes such as inspection, gouging, welding, grinding, and polishing. Demonstrations in real-world production settings, especially in critical sectors such as automotive and aerospace, aim to showcase the technology's potential. The market opportunity for these self-configuring robotic processes is estimated to be around 2000 robotic workcells, representing a market value of approximately 600 million euros.