Human-Centered Holographic Assistant Using Computer Vision to Interact with Battery Manufacturing Machinery in a Predictive Manner

This work focuses on a novel human-centered digital assistant combining Mixed Reality (MR), Computer Vision and Machine Learning regression to guide professionals and students on how to operate and correctly parameterize battery manufacturing machinery. Our Concept article aim is to provide a proof of concept of our digital assistant, for a process involving an operator interacting with a semi-manual electrode calendering machine and looking how the intended calendering parameters will impact the electrode properties after calendering. The operator performs his/her actions while our digital assistant automatically detects the parameters entered by him/her on the machinery. Then, our digital assistant provides real-time predictions to the operator, helping him/her in decision-making, through a minimalistic holographic interface minimizing visual hindrance. The ergonomics of our solution was optimized by acquiring feedback from various experience levels users evaluating their performances. As the necessity for advanced energy storage solutions rises, there is a strong need for modern training and guidance tools that aid in complex battery manufacturing processes at the prototyping stage involving both semi-manual and automatic activities. Our digital assistant has potential to improve manufactured electrode and cell qualities by placing emphasis, thanks to MR, on the egocentric perspective along the battery cell prototyping process.