Autonomous Rover Navigation on Unknown Terrains: Functions and Integration

Abstract

Autonomous long-range navigation in partially known planetarylike terrains is still an open challenge for robotics. Navigating hundreds of meters without any human intervention requires the robot to be able to build various representations of its environment, to plan and execute trajectories according to the kind of terrain traversed, to control its motions and to localize itself as it moves. All these activities have to be scheduled, triggered, controlled and interrupted according to the rover context. In this paper, we briefly review some functionalities that have been developed in our laboratory, and implemented on board the Marsokhod model robot, Lama. We then present how the various concurrent instances of the perception, localization and motion generation functionalities are integrated. Experimental results illustrate the functionalities throughout the paper.

Robust Autonomous Navigation and World Representation in Outdoor Environments

“…A sparse representation formed only by isolated landmarks will in general not fulfil the necessities of an autonomous vehicle, and a more detailed representation will be needed for tasks such as place recognition or path planning. Furthermore, not only is a dense representation of the environment required, but also an algorithm that is able to obtain multi-layered maps, where each layer represents a different property of the environment, such as occupancy, traversability, elevation, etc (Lacroix et al, 2002).…”

Probabilistic scan matching for motion estimation in unstructured environments

“…In robotics, they have been successfully used in a wide range of applications as an ameliorated odometry (i.e. simultaneous location and map building [20], [8], [10], local map construction [14], [11], [15] or people tracking systems [19]). …”

Monte Carlo Localization in Outdoor Terrains Using Multi-Level Surface Maps

“…Additionally, they apply a scan-matching routine for localization, which does not facilitate global localization. To reduce the memory requirements of outdoor terrain representations, several researchers applied elevation maps (Bares et al, 1989;Hebert et al, 1989;Lacroix et al, 2002;Parra et al, 1999). A probabilistic approach to localize a planetary rover in such elevation maps has been described by Olson (2000).…”

“…In this paper, we consider the problem of localizing a mobile robot in outdoor environments by matching laser range measurements to a given map of the environment. One of the most popular representations for outdoor environments are elevation maps (Bares et al, 1989;Hebert et al, 1989;Lacroix et al, 2002;Parra et al, 1999). The key idea underlying elevation maps is to store the height information of the terrain in a two-dimensional grid, which corresponds to a representation of the horizontal surfaces of the environment.…”