Plowing for Controlled Steep Crater Descents Jason Ziglar, David Kohanbash, David Wettergreen, and William Whittaker Carnegie Mellon University jpz@cmu.edu, dkohanba@cs.cmu.edu Abstract Controlling slip is crucial to reliable descent of steep un- consolidated slopes such as those found in lunar craters. The ability to enter these craters provides opportunities to explore potential in-situ resources, such as water ice. We present a rover prototype with a novel, actuated, omni-directional plowing device and control method for maneuvering on steep slopes. Data from field experiments show reliable control dur- ing descent on loose sand slopes up to 40 ◦ , with twenty-fold reduction in downhill slip and a threefold reduction in slip during point turns. In particular, the data indicates that plow- ing eliminates slip caused by shear failure created in angle-of- repose material. 1. Introduction Discovering water ice and other important volatiles on the lunar surface drives attempts to explore the interior of lunar craters. Volatile trapping craters may range in size from a few tens of meters to tens of kilometers in diameter. Plans for extended human presence on the lunar surface benefit sig- nificantly from in-situ resource utilization [14]. With evidence that cold traps located in craters at the lunar poles potentially contain water ice [1, 8, 5, 20] controlled descent of crater slopes is a keystone to exploring these resources. Models for lunar craters estimate unconsolidated regolith slopes with a 30 - 40 ◦ angle-of-repose [6, 7]. These slopes ex- hibit fluid-like flows of soil, resulting in an uncontrolled “surf- ing” descent. Theory suggests that for loose, granular soil, strength lies under the surface, not on, the surface [18, 21]. In order to explore rover technologies for this regime of locomo- tion, Icebreaker, a prototype rover is presented. The primary objective of this rover is to explore and develop concepts of locomotion for crater descent. Research pushes technologies and configuration requirements for planetary rovers intended to descend steep crater walls. The primary innovation is the use of plowing to control descent. 2. Rover Design Goals of steep crater navigation come from a mission pro- posal to send a small, low cost rover as a secondary payload opportunity [3]. The mission framework called for exploration of lunar craters with little available space; a compact, capa- ble rover could fit the requirements, but costs would require a rover capable of landing outside a crater to descend inwards. These limitations drive the design towards the current form. For climbing, low center of gravity prevents tipping on steep slopes, high flotation maximizes locomotion in loose soil, and low ground pressure limits the rover sinkage. Tracked vehicles exhibit advantages for these characteristics. Tracked vehicle stance can be wide, long, and low for stability in steep terrain. Components like batteries and computers can be placed within the tracks, keeping mass close to the ground. Tracks can grip terrain and bridge irregularities with high flota- tion. Increased surface area along the tracks spreads weight to keep ground pressure low. As a result, the Icebreaker rover is capable of the steep and steady descent and features a low center of gravity, high traction, and low ground pressure. Figure 1. Icebreaker chassis design. (i) Internal volume for components (ii) Tracks for locomotion. The rover chassis, as seen in Figure 1, shows the chassis design of the rover. The rover measures 1.4 meters long, 1.1 meters wide, and 0.3 meters tall. The chassis provides a rigid frame to which tracked side-frames are attached. The frame maintains the simplicity of design, reducing weight and com- plications arising from an articulated or actuated frame. The side-frames provide an internal volume to contain components for robot operation. Instead of idler wheels along the length of the track, Teflon guide bars provide low-friction support with- 1