Obstacle avoidance and navigation in robotic systems: A land and aerial robots study

Autonomous airborne systems have generated a lot of interest in civilian and military applications. The operation of such systems involves routing and navigation toward targets and obstacle avoidance. In this paper, these problems were tackled and solutions were applied to land-based robots as well as a quad-rotor aerial system. The quad-rotor flight path navigation and routing was programmed based on GPS and on-board measurement data. Obstacle avoidance was implemented based on an algorithm that relies on the idea of Virtual Potential field. Kalman filters were implemented to improve the accuracy of the measured data. While 3D visualization was used to visually identify obstacles. In the case of in-building reconnaissance, where GPS signals are very weak and largely useless, we rely on laser sensors and data aggregation from proximity feeds to identify obstacles and the shape of the surrounding environment.