Advancing packaging at ALUMIL: ALUMIL's progress with the SMARTHANDLE Project
As part of the SMARTHANDLE project, ALUMIL is developing a dual-arm robotic packaging solution for handling large, multi-variant aluminium profiles. In collaboration with the University of Patras, this pilot aims to boost packaging efficiency, reduce errors, and improve worker safety, while paving the way for smarter, more flexible manufacturing.

Motion planning with geometric constraints
UPC has developed a motion planning framework within SMARTHANDLE to enable flexible, constraint-based robot movements for industrial tasks. Built on OMPL and integrated into the Kautham platform, it allows easy configuration, 3D visualization, and real-world testing, enhancing robot adaptability in scenarios like assembly and manipulation.

From proof-of-concept to full prototype: Smart handling in automated manufacturing
The SMARTHANDLE project has delivered custom prototypes for automating Menicon’s contact lens production, including a smart gripper with tactile feedback and a flexible transport system. Developed with Demcon and STT Products, these solutions are now entering real-world testing ahead of integration into the pilot production line.

Dual-arm robot system for aluminium profile handling
KUKA is contributing to the SMARTHANDLE project by developing a dual-arm robotic system for automated aluminium profile handling. The setup features advanced 6D pose estimation, AI-driven grasp selection, and collision-free motion planning, supported by impedance control for safe operation. Initial tests were successful, and future improvements will focus on enhancing pose accuracy, integrating smarter grasp planning, and adopting adaptive, object-based path execution.

Hand gesture recognition as a quick interface to command robot behaviour
The SMARTHANDLE project is developing a hand gesture recognition system that uses machine learning and real-time visual processing to control robots intuitively through natural movements. Currently being developed at AIMEN for use with different robot models, the system is being refined for greater accuracy and user-friendliness by training on diverse hand data and simplifying gesture execution.

AI-based 6D Pose Estimation Leveraging the Power of Synthetic Data
AI-driven 6D pose estimation is transforming industrial automation by enhancing robotic vision and efficiency. Synthetic data provides a scalable solution for training deep learning models, addressing challenges in real-world data collection. A case study in the metal industry showcases how Roboception leverages Blender-generated synthetic datasets to improve object detection and pose estimation, ensuring adaptability in dynamic manufacturing environments.