General information
Reference : UMR7253-PHIXU-007
Workplace : COMPIEGNE
Date of publication : Monday, May 9, 2022
Scientific Responsible name : Philippe XU
Type of Contract : PhD Student contract / Thesis offer
Contract Period : 36 months
Start date of the thesis : 1 November 2022
Proportion of work : Full time
Remuneration : 2 135,00 € gross monthly
Description of the thesis topic
Intelligent navigation systems for fully or partially delegated vehicles, as well as driver assistance systems, require an understanding of the driving environment with a high level of integrity. A lot of navigation information is stored in road maps and is used in real time with the help of a localization system. Like any source of information, maps are subject to errors, which can sometimes be made during the mapping phase, but are more often due to a change in the road environment itself. The use of these maps in a navigation system requires them to be constantly monitored. Perception systems, by cameras or lidar, allow to check the integrity of the map information by a real-time vision of the navigation environment. The problematic of this thesis is to study the joint use of a map and an on-board localization and perception system in order to quantify the quality of the map information. The main issue is the definition of metrics to qualify the usability of map information in navigation tasks. Since perception systems are also subject to errors, information redundancy is necessary to guarantee a sufficiently low level of risk. This redundancy can be achieved through multiple trips or through a fleet of collaborative vehicles.
Work Context
This thesis subject is part of the joint laboratory SIVALab between Renault and Heudiasyc. On the Renault side, two research engineers will be involved in supervising the work. They are Marie-Anne Mittet and Eric Gosselin, respectively "localization" and "cartography" experts at Renault's Technocentre.
We talk about it on Twitter!