Robustness Concepts for Hopping Robots

Abstract

This paper presents a few concepts for the robustness of the paradigm for mobile robots. We are motivated to develop these concepts by the need to make realistic advances toward smaller planetary rovers and to allow more frequent, and possibly cheaper, space missions to small planetary bodies. Small celestial bodies are usually characterized by a low to medium gravitational environment and unstructured terrain, thus hopping mobility has certain advantages v/s more traditional mobility methods. The work described in this paper arises from an obvious debate about the best way to design exploratory robotic systems under the very severe constraints on launch mass. For a given amount of launch mass (100 kg for example), is it preferable to have one or two exploratory vehicles (weighing 50 or 100 kg in this example), or many small vehicles (e.g., fifty 2-kg mass vehicles), or one large vehicle with many small helpers (e.g., a 50 kg vehicle with 25 2-kg assistants)? To date, the Pathfinder mission to Mars and the Mars exploration missions planned for the next decade have focused on the single vehicle paradigm. The only successfully deployed mobility paradigm for autonomous exploration of planetary surfaces is a 6-wheeled rover, as seen in the Pathfinder mission's Sojourner vehicle [1], and in proposed Mars exploration missions planned for the next decade. Because of its unique rocker-bogey suspension, a 6-wheeled rover of the Sojourner type can traverse obstacles that are about 1.5 times the vehicle's wheel diameter. However, this still represents only a fraction of the vehicle's overall body length. Wheeled designs have fundamental limitations on the obstacle size, compared to body length, that can be overcome. Terrain accessibility may become a problem as vehicles are scaled down in size in order to enable multivehicle design approaches, for example the Jet Propulsion Laboratory – NASA nano-rover is a small (~1 kg mass) 4wheeled exploratory rover [2] that can only go over obstacles a few cm in height, because of the fundamental limitations imposed by wheels. Legged robots can overcome the limited traversability of wheeled vehicles in many rugged terrains. Legged rovers have previously been proposed for Lunar and Martian exploration [3], and large legged vehicles have been demonstrated in the tough environment of an Alaskan volcano [4].While legged robots can potentially access rough terrains, they are mechanically complex, requiring numerous joints, actuators, and linkages. While spiders and insects can demonstrate impressive ability to climb over obstacles, it is not yet clear that multi-legged robot vehicles have impressive ability to overcome large obstacles when the are scaled down in size. Clearly, small vehicles operating in complex terrains may have to overcome obstacles which are equal to, or larger than, the vehicle's size. Vehicles that can jump or hop as part of their operation might be able to overcome such relatively large obstacles if they can reliably survive the hard landings inherent to the jumping process. Furthermore, as the magnitude of gravity is reduced, the practical efficiency of hopping increases relative to wheeled vehicles. Hence, hopping can be a realistic alternative to wheels in lower gravity environments. These simple observations motivate the use of leaping motions for small robots in unstructured terrains. Our work suggests that hopping robots may be a useful addition to the planetary rover family (e.g., they may operate in tandem with conventional rovers). They may also be suited for the cooperating behavior planned for the next phases of Mars exploration, wherein many simple devices will coordinate their motions to collectively gather distributed scientific data over large areas. All of this, naturally, provided that the robot is strong enough to survive repeated crash landings on rocky terrain. For this reason, the paper proposes an approach to landing protection based on a new concept of air bag