Skip to Main content Skip to Navigation
Theses

Conception d'une interface avec réalité augmentée pour la conduite automobile autonome

Abstract : This doctoral thesis was conducted under the canopy of the Localization and Augmented Reality (LAR) project. The research project was focused on railway yard and autonomous vehicles. The thesis provides answers to three main questions about the Human-Machine interface design in autonomous vehicles: Which information should be conveyed to the human agent? In which form? And when? Answers will enable an appropriate trust calibration of the human agent in the autonomous vehicle and improve driver’s experience by making automation “transparent”. We focus especially on the lane changing task entirely realized by the autonomous vehicle. The aim and the objectives were achieved by a five-steps methodology. Some general principles of transparency have been redefined on the LYONS (2013) model. These principles have been then operationalized by means of Cognitive Work Analysis. Graphical representation of useful information or potentially useful information was defined during creative sessions, by using Augmented Reality that lies at the heart of the LAR project. This information was categorized according to the functions from which it results: information acquisition, information analysis, decision making and action execution. Five interfaces were designed. Each of these interfaces presents information from some of these functions. Therefore, these interfaces corresponded to different transparency configurations more or less high. The validity of transparency principles was tested through an experiment on driving simulator with a sample of 45 participants. In this experiment, some indicators of cognitive activities and User Experience were measured. Data analysis has led to some differences between 5 interfaces. Indeed, the interface with related information to “information acquisition” and “action execution” functions improves the cognitive activities of the human agent. Considering the User Experience, it is the interface with the information related from the 4 functions which provides the best User Experience.
Complete list of metadatas

Cited literature [447 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02096248
Contributor : Abes Star :  Contact
Submitted on : Thursday, April 11, 2019 - 11:09:06 AM
Last modification on : Saturday, February 29, 2020 - 3:41:52 AM

File

Pokam_Meguia_Raissa2.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02096248, version 1

Citation

Raïssa Pokam Meguia. Conception d'une interface avec réalité augmentée pour la conduite automobile autonome. Automatique / Robotique. Université de Valenciennes et du Hainaut-Cambresis, 2018. Français. ⟨NNT : 2018VALE0029⟩. ⟨tel-02096248⟩

Share

Metrics

Record views

365

Files downloads

256