Authors:
Henrik Schiøler
;
Luminita Totu
;
Anders la Cour-Harbo
;
John Josef Leth
and
Jesper Larsen
Affiliation:
Aalborg University, Denmark
Keyword(s):
UAS, 3D Mapping, Indoor Path Planning.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Formal Methods
;
Industrial Automation and Robotics
;
Industrial Engineering
;
Informatics in Control, Automation and Robotics
;
Intelligent Control Systems and Optimization
;
Intelligent Transportation Technologies and Systems
;
Mobile Robots and Autonomous Systems
;
Planning and Scheduling
;
Robotics and Automation
;
Simulation and Modeling
;
Symbolic Systems
Abstract:
Indoor operation of micro Unmanned Aerial Vehicles (UAV or UAS) is significantly simplified with the capability for indoor localization as well as a sufficiently precise 3D map of the facility. Creation of 3D maps based on the available architectural information should on the one hand provide a map of sufficient precision and on the other limit complexity to a manageable level. This paper presents a box based approach for easy generation of 3D maps to serve as the basis for indoor navigation of UAS. The basic building block employed is a 3D axis parallel box (APB). Unions of APBs constitute maps, which are by definition closed to basic set operations such as; union, intersection, set difference etc. The restriction to APBs is made in accordance with the tradeoff between simplicity and expressiveness, where real time requirements emphasize simplicity. The mapping approach is presented along with different approaches of selecting via points ensuring sufficiently efficient path plannin
g and at the same time ensuring scalability by keeping complexity low. A proposition for {\em minimal} via point assignment is ensuring at least feasibility of path planning is presented. Feasibility is not proved formally. Instead results from a randomized {\em statistical proof} are given.
%An easy-to-use procedure for calibrating maps and indoor positioning is presented.
3D Mapping, via point assignment and calibration are all implemented in a stand alone software application. The application program provides, through a graphical user interface, the possibility to map indoor environments based on existing 2D floor maps.
(More)