Date of Award

8-31-2022

Publication Type

Thesis

Degree Name

M.A.Sc.

Department

Mechanical, Automotive, and Materials Engineering

Keywords

LiDAR;Mobile Robot;Navigation;ROS;Sensor;Specular Reflection

Supervisor

Jalal Ahamed

Abstract

Simultaneous Localization and Mapping (SLAM) is the process of representing the spatial environment (mapping) while keeping track of position (localization) within the built map. SLAM is widely used in indoor navigation, where several types of objects and obstacles need to be mapped. Objects that pose an issue for laser or light-assisted indoor navigation include specular surfaces such as mirrors that cause light reflection. This thesis aims to understand the characteristics of mirror-like objects in various arrangements. Experiments were conducted using a lidar-mounted mobile robot navigating with respect to one or more mirrors and mapping the environment. Several observations were made to understand the inaccuracies of laser scans in mirrored environments. The outcomes of these observations suggest that laser scans may be fully reflected off mirrors, causing no range or intensity data to be provided back to the robot and causing the map to develop areas of negative space. Objects or boundaries within the range of the lidar may be mapped behind the surface of the mirror, and self-detection may occur on the surface of the mirror. Some uncertainties may occur when more than one mirror is present in the environment. As many different observations were noted, a potential solution approach is outlined, advising the use of clustering algorithms to identify and remove inconsistencies before building a map of an indoor environment. This research has applications in several industries. These could include autonomous robots navigating through the environment to perform specific tasks; mapping of malls, museums or office buildings where specular surfaces are common. Future work would include implementation of the solution approach and extension to different types of nondiffuse surfaces, including transparent and translucent surfaces. This research could be most practical when one or more mirrors are present. applied in the mapping of indoor spaces with high amounts of optically unique surfaces, including modern office buildings, hospitals, and museums

Download

Included in

Robotics Commons

Share

COinS