There is a strong trend to automate the manufacturing industry and investments into automation are ongoing. Classical automation works well for big producers with large series of products. However, it reaches its limit by smaller series, specialized industries and SMEs. These cannot afford replacing manual workers with robot programmers who cost about two times more, who lack the knowledge of the specialized production technology and therefore need too much time and resources to fine tune the robotic programs.

This created a demand for no-code solutions for robot programming based on digital twins, technology simulations, use of virtual reality and generative solutions based on production data. However, these solutions are often implemented in a rigid way, complicated to use for ordinary workers, unexplainable and therefore untrustworthy and unreliable.

The Challenge 4 ‘Digitalization procedure for production tools &  machines for industry 5.0’ states clearly that in order to ensure the shift on the shop floor it is needed to implement digitalization with care. Attention must be paid to the most valuable resource in production – human skills and know-how, experience and critical thinking. Digitalization of production must be human centric and respectful in order to preserve benefits of this resource.

Therefore, we propose Manuvised – a solution for automation of manufacturing tasks that learns from human demonstrations, allows workers to supervise it and takes in feedback from this supervision to immediately correct imperfections and learn from it in long-term. Manuvised changes manual routine work into collaborative automation.

RoboTwin’s long term focus is easy robot teaching. Our experience with the industry show that manufacturers need a reliable, trustworthy solution, integrated into a production system, understandable rather than a black box, flexible to adapt to changing product portfolio and easy to use by ordinary workers. Manuvised is our answer to this need. Thanks to IKERLAN’s support via their building block 5 ‘Robot Intelligent Control’ and collaboration, RoboTwin will benefit from IKERLAN’s expertise in the field of real-time control of robots (KUKA, UR or other), but also integration of deep learning control algorithms for static manipulators in industrial HW and integration of HW peripheral devices into abased system.