Autonomous cross-country navigation with the ALV

Abstract

This paper describes the first cross-country map and sensorbased autonomous operation of a robotic vehicle. Experiments on the Autonomous Land Vehicle in natural terrain were performed. An overview of the software architecture used for this achievement is discussed, and details of the perception and planning techniques are presented. We describe two key experiments where the vehicle avoided known and unknown obstacles in its path.

Vision for mobile robot navigation: a survey

“…In those cases, a map of the traversed areas has to be created and a localization algorithm has to be implemented in order to carry out goal-driven navigation. When maps need to be built by a robot, they may be created in either a vehiclecentered coordinate frame vehicle [81] (and updated [82] as the vehicle moves) or with respect to some external reference, such as an external camera attached to a companion device [99] (as in Mars Pathfinder and Lander), or a virtual viewpoint with respect to which range data is projected to form a Cartesian Elevation Map [30] or a global positioning reference such as the sun [29].…”

Real-Time Photorealistic Virtualized Reality Interface for Remote Mobile Robot Control

“…Recent outdoor autonomous vehicles compile volumetric representations (Lacase, Murphy, & DelGiorno, 2002) before searching for the supporting surface (Wellington & Stentz, 2004). Methods for predicting motion under a terrain-following constraint are fundamental to obstacle avoidance and they have been used since the first outdoor mobile robots (Daily, et al, 1988).…”

Toward Optimal Sampling in the Space of Paths

“…This has the effect of guiding the vehicle towards x = ∞. The paths are each of length 17 m. The method of evaluating a constant set of paths has been applied to the obstacle avoidance problem at least as far back as [4].…”