Active airborne localisation and exploration in unknown environments using inertial SLAM

Future unmanned aerial vehicle (UAV) applications will require high-accuracy localisation in environments in which navigation infrastructure such as the Global Positioning System (GPS) and prior terrain maps may be unavailable or unreliable. In these applications, long-term operation requires the vehicle to build up a spatial map of the environment while simultaneously localising itself within the map, a task known as simultaneous localisation and mapping (SLAM). In the first part of this paper we present an architecture for performing inertial-sensor based SLAM on an aerial vehicle. We demonstrate an on-line path planning scheme that intelligently plans the vehicle's trajectory while exploring unknown terrain in order to maximise the quality of both the resulting SLAM map and localisation estimates necessary for the autonomous control of the UAV. Two important performance properties and their relationship to the dynamic motion and path planning systems on-board the UAV are analysed. Firstly we analyse information-based measures such as entropy. Secondly we perform an observability analysis of inertial SLAM by recasting the algorithms into an indirect error model form. Qualitative knowledge gained from the observability analysis is used to assist in the design of an information-based trajectory planner for the UAV. Results of the online path planning algorithm are presented using a high-fidelity 6-DoF simulation of a UAV during a simulated navigation and mapping task.