Perception Testbed and Lighting

Figure 1: Testbed for evaluating MoonRanger’s perception. Material, color and lighting are lunar-like. The line of rectangular blocks are at known locations for verifying that the terrain geometry created by the robot matches the geometry of the scene.

LUNAR POLAR LIGHT

Figure 2: The wavelength distribution and intensity of light that reaches the Moon’s surface (orange envelope) exceeds the less intense and complete color spectrum that reaches the Earth’s surface (colorized envelope).

It is challenging to illuminate a perception testbed with lighting that emulates sunlight that the rover will experience on the pole of the Moon. This is not nearly as simple as going outdoors on Earth. The Moon’s lighting is more intense, less-diffuse, and lower in the sky than what we experience on Earth. The light reaching the Moon is 1350 watts/meters2. By comparison, the light reaching the Earth is only 1120 watts/meters2. The Moon’s light is direct from the sun and has full infrared, but the Earth’s atmosphere reduces infrared (see the orange envelope of lunar sunlight versus the colorized envelope of Earth’s sunlight that appears in Figure 2). The atmosphere also bounces the light as if coming from all directions, not just straight from the sun. For this reason Earth shadows are not as dark or sharp as Moon shadows. The sun never rises more than a few degrees at the Moon’s pole, so when at the pole the sun sits on the Moon’s horizon, and incoming light is nearly tangential.

EMULATING LUNAR POLAR LIGHT

Figure 3: Arrimax cinematic day light (right) casting 18 kilowatts of lunar-like illumination onto MoonRanger’s perception testbed.
Sun power and sun spectrum as experienced at the lunar pole is needed for testing MoonRanger’s cameras, lasers and perception. A special light for movie-making does just that. This uses a Hydrargyrum Medium-Arc Iodide (HMI) bulb. An intense 18 kilowatts of power impinge on the scene. Blackout curtains prevented light from reflecting diffusely back onto the terrain.