X PRIZE Foundation
Login / Register
Automotive X PRIZEArchon X PRIZE for GenomicsSpace
Spanish
Russian
French
German
Chinese
Korean
Japanese
Portugese
Italian

Progress


<< previous      |      >>

My newly purchased HDTV camera actually weights in at 0.4 grams! It has only a simple lens, not a Zoom model. For the GLXP imaging it will be necessary to add a small “Converter” lens unit, particularly for telephoto operation (0.3 milliradian per pixel required, or about 3x telephoto effect.) This small lens unit could be flipped over to produce a wide angle mode, but it is easier to slide in a wide angle or “Fish Eye” adapter. In fact, these cameras are so low mass – compared to any reliable mechanical mechanism – that it is better to provide three separate cameras, two with the special optics attached. Providing three cameras also makes video capture possible – with reduced options – if one camera stops working – nice insurance when $$ Millions have been invested in the flight!

The telephoto optical system will need adjustable focus, for the depth of field at this magnification becomes quite limited. No aperture adjustment is necessary as the gain control modes on the camera are sufficient. I envision the Lander “Self Portrait” actually using the adjustable focus telephoto system, and a convex mirror. A very light convex mirror, held away from the lander on a long arm, could reflect an image of the entire lander. Viewed by the telephoto optics, this could provide a usable “Portrait”. Computer image correction could remove distortions and yield a good photo. Alternatively, an additional camera unit could be mounted on that “arm” and be lighter than a mirror.

The Lunar Navigation “Cameras” I proposed in one of our SBIR submissions, may be desirable to quantify the location of our lander, when it completes its 5km “Hop”. Several companies are working on “HDR” (High Dynamic Range) video cameras for automotive and other uses. The military has been working with high sensitivity versions of these for over a decade. In low light, the “High Dynamic Range” keeps bright lights from masking important details in the glare. In daylight, glare from the sun produces similar problems. Photographers quickly learn to avoid these problems, but “full time” cameras – for continuous monitoring – encounter these conditions with some regularity.

One recently advertised commercial unit promises 110 dB dynamic range: a brightness ratio of 300,000:1. This makes it possible to image the naked sun, and sun lit Earth or Moon surface details in the same image. (Roughly a 40,000 brightness ratio, Sun's surface brightness, to white target on the Earth.) At night, it makes it possible to image both bright stars and the sun lit Earth in the sky. These combinations are often required for Lunar navigation (if Lunar expeditions will continue after sundown and on the “Far Side”.) The worst case situation will still require use of a “Sun Shade”, but that is not a serious problem when accurate navigation is needed to correct for accumulated errors in “Dead Reckoning”. This extreme occurs in daylight on the Far Side, when only the sun and stars are visible, and the sun – a single point reference – provides a good direction, but a second reference is necessary to compute an absolute position.


Comments

Rover navigation

It will be very interesting to see the results of various teams' approaches to rover location and navigation as they play out over the coming years. You mention dead reckoning, which has obvious difficulties, especially over long distances, and especially due to wheel slippage on slopes. The idea of establishing location by imaging celestial objects is well established on Earth, and it will be interesting to see it put onto practice on the Moon. I would be concerned about the level of accuracy that can be achieved.

However, assuming it's very good, we also have to deal with the issue of map accuracy. Location from sky observation gives coordinates, but if the nearby surface features are not placed accurately on the map, the coordinates don't help lead us to a desired location among those features. Errors of 1000 m or more are not uncommon on maps of the near side. They will be improved by LRO, but not before the first GLXP teams fly if they meet their goals.

Another approach to navigation is to fix location by comparison with surface features. First, triangulation between several surrounding features locates the rover on an orbital image. Then the rover proceeds via defined waypoints - drive to this crater rim, turn east and proceed to that hill, at its foot, turn right and drive to the small fresh crater - and so on. This is in effect what the Mars rovers do. I see it as a more effective near term solution, while more sophisticated systems and better map controls are developed. We already have all the images we need for hundreds of great landing areas.

Phil Stooke
Author of "The International Atlas of Lunar Exploration"
Departments of Geography and Physics/Astronomy
University of Western Ontario, London, Ontario, Canada