PLANETARY EXPLORATION USING HOPPING ROBOTS

摘要

The recent trend towards small and frequent space mission to Mars and other celestial bodies such as moons, asteroids, and comets has sparked new interest towards multi-functional vehicles, capable of providing excellent mobility to dedicated scientific packages. Currently, the only deployed, and actively engineered, mobility paradigm is a 6-wheeled rover, as seen in the Pathfinder mission's Sojourner vehicle [12] and in the planned Mars 2003 exploration missions. Most 6-wheeled rover designs can traverse obstacles that are about 1.5 times their wheel diameter, but they also have significant drawbacks preventing their use as truly general exploration platform. For example, they can only drive over obstacles that are a fraction of the vehicle's body length, and use a significant number of actuators and complex suspension linkages. To address the two goals of reducing rover complexity and developing more efficient mobility methods, in the past few years we have been developing planetary robots equipped with a very small number of actuators and capable of moving by hopping. Since planetary bodies of current interest are characterized by low to medium gravitational environments, wheeled mobility can be replaced, for certain operations, by hopping, which is a more efficient mobility method. Key to the good performance of these systems is the trade-offs between functionality and complexity in the context of the design and development of a small robot, capable of moving a camera and a science package by jumping. Our hopper's operation is more akin to the movement of a frog, rather than the oscillatory behaviour of typical hopping robots [14]. In particular, with our first prototype we have investigated the basic problems of rover jumping. With the second prototype we have shown that a system weighting less than 1.5 kg can efficiently convert the energy stored a single actuator to propel, steer, self-right a simple hopper, and pan an on-board camera. With the third prototype, we have shown the ability to combine coarse motion generated by hopping with wheeled motion for fine position of science instruments. Figure 1 shows a picture of the first three generations of hopping robotis. There are, however, several important aspects, which must be solved before this mobility concept can be used in real exploration missions.