Environments characterizing disaster zones frequently impose locomotive
limitations on traditionally larger, wheeled or tracked mobile platforms
due to the inability to drive over large obstacles and a lack of
precise maneuverability. Alternative robotic solutions demonstrate
potential to overcome many of these limitations; snake-like robots, in
particular, are advantaged in narrow, confined spaces, desert-like
terrain as well as arboreal environments, among others that may be
encountered in disaster zones comprising fallen rubble and blocked
paths. Larger, wheeled robots, on the other hand, possess high-speed
mobility useful for expedient traversal of large, obstacle-free expanses
of terrain. The latter additionally support greater carrying capacity,
allowing for more capable on-board computing as well as power capacity.
We integrate instances of each class of mobile robots into a marsupial
robotic pair that, together, leverages their individual strengths while
limiting the drawbacks of either, to accomplish autonomous exploration
in a planar, obstacle-cluttered environment. The marsupial pair, and
accompanying robotic architecture we develop, are challenged to explore
and navigate an arbitrary obstacle configuration in a planar indoor
environment. In a collaborative manner that utilizes each platform
according to its strengths, the Turtlebot and robotic snake employ SLAM
and frontier exploration algorithms, in conjunction with locomotive
primitives available to each, to cooperatively explore, map, and
navigate an initially unknown scenario.
Menu