Thesis defense by Quentin LOIZEAU

Due to circumstances, this defense will be conducted via teleconference using the Teams platform. You can connect to it using the following link:Click here to join the meeting

Summary

Augmented reality (AR) is gradually being deployed in industry for very specific tasks. This deployment is hampered by a lack of information on the real added value of AR. Previous work highlights its value for complex tasks where humans interact with large amounts of information. However, the added value of AR in an industrial context is difficult to quantify for these tasks, for which it is not possible to transpose results observed in the laboratory, and experiments require significant resources. We propose a data model that characterizes complex maintenance tasks and takes into account the impact of AR on performance and users for training classification algorithms to predict this impact before deployment. We are conducting field experiments to validate the selected evaluation criteria and collect data to feed into the classification process. This work is leading to the selection of the most relevant characteristics for predicting the added value of AR on complex maintenance tasks.

Jury

Ms. Saida BOUAKAZ, Professor, LIRIS – UMR 5205, University of Lyon 1 - Chair

Mr. Florent LAROCHE, Doctor of Philosophy, LS2N – UMR 6004, Ecole Centrale de Nantes - Rapporteur

Mr. Frédéric NOËL, Professor, Industrial Engineering Department, Grenoble INP - Rapporteur

Mr. Frédéric MERIENNE, Professor, LISPEN – EA7515, ENSAM - Examiner

Ms. Florence DANGLADE, Professor, LISPEN – EA7515, ENSAM - Examiner

Mr. Fakhreddine ABABSA, Professor, LISPEN – EA7515, ENSAM - Examiner

Mr. Francis ROSSIGNOL, VP Smart Factory, Safran Landing Systems - Guest