Etude exploratoire d'un assistant digital d'aide au montage basé sur la projection en réalité augmentée: An Exploratory Study of a Projection-based Augmented Reality Digital Assistant to Facilitate Assembly.

This article presents PMF-Vision, a digital assistant for manual assembly based on the concept of augmented reality information projected on the workspace. This digital assistant aims at assisting operators with mild mental deficiencies in their assembly tasks by tracking their actions in real-time. In this research, we analyzed hand tracking algorithms based on the latest deep learning techniques and highlighted the best candidate for our context in terms of accuracy and processing time. We present different interaction concepts for the system and propose to evaluate two visualisation concepts to assist the operator during component placement. The results of the exploratory study (N=6) highlighted the excellent usability of the system and the interest, expressed by the target population, to use such tools in their daily tasks.