The heart of autonomous warehouses lies in their stacker cranes, and any unplanned downtime translates into inaccessible goods and delayed deliveries. Our approach involves a monitoring system complemented by models that predict potential failures in advance. This system enables scheduled maintenance, leading to a reduction in unplanned downtime, and increased reliability in the operations of autonomous warehouses.

Addressing the fragmentation in robot programming and integration ecosystems, this project aims to develop a universal software capable of controlling any industrial robot in real-time. It encapsulates complexity, fosters solution reusability, and democratizes robot programming, making it accessible and efficient.

Predicting the temperature field evolution during Wire Arc Additive Manufacturing (WAAM) with short computational times is essential for optimizing the process. This work introduces an innovative numerical scheme for fast thermal simulation for WAAM, combining process knowledge and high performance computing. The computational time is shorter by a factor of 10 compared to the simulated manufacturing time, allowing integration in iterative planning.

Producing new workpieces with Wire Arc Additive Manufacturing (WAAM) First-Time-Right is challenged by the uncertainty of the process behaviour. This work developed the basis for a model process planning approach for WAAM. The resulting framework is composed by the models, control approaches and system requirements needed for a successful first time right approach.

This work addresses external forces and robot model inaccuracies by utilizing a Creaform optical tracker for real-time pose tracking. A novel prediction technique, leveraging joint state information, mitigates high measurement delay. Impressive outcomes feature a 50μm mean path error at high welding speeds and compensation for external forces up to 5Hz, exceeding the robot's rated payload.