CMU / NASA Partnership Advances Research in Lunar Topography-Silicon Valley Campus - Carnegie Mellon University

CMU / NASA Partnership Advances Research in Lunar Topography-Silicon Valley Campus - Carnegie Mellon University

CMU / NASA Partnership Advances Research in Lunar Topography

New research from NASA (National Aeronautic and Space Administration) is producing the most accurate topographical maps of the lunar surface yet. By revisiting previous flights and their telemetric data with new mathematical processing algorithms, the results of those explorations can be refined to become significantly more accurate. Carnegie Mellon University Silicon Valley students have brought their programming expertise to provide the means to turn the research into real results. Initial testing on the Apollo 17 mission have shown dramatic improvements in the usability of the data, and this research is expected to be useful for all cartographic related missions.

One of the main purposes of NASA missions is cartography of planetary bodies, but unfortunately the original flight data is not always accurate. In the case of the Apollo missions, which is the case study under research, they have particularly noisy data sets, and the cartographic photography is therefore of limited use. Noise in the electronics provided erroneous data points in terms of location and time of each photograph, thus images taken from this flight cannot be accurately mapped to the lunar surface. Mathematician Dr. Taemin Kim proposed an algorithm of refining the telemetric using Conjugate Gradient methods to predict more accurate results for each data point by minimizing the calculated error, and then translating each image using its shifted telemetry. This process of refining the telemetric data is known as Orbital Refinement, and applying these results to cartographic images is known as Bundle Adjustment.

Computer scientist Zachary Moratto from NASA teamed up with CMU students Darryl Melander, Henry Fung, and Daniel Hebb, to provide a software implementation for this research effort. Building on NASA’s open source product, Vision Workbench, the joint team has added functionality to perform both Orbital Refinement and Bundle Adjustment using this new Conjugate Gradient linear solver in the algorithms. This software will be refined by computer scientists at NASA to be released as open source software under the NASA Open Source Software Agreement.

Working very closely together, the team has analyzed data from the Apollo 17 mission. The results of Dr. Kim’s research have proven to be a great improvement over the current data set. NASA has generously provided an updated topographic model of the lunar surface, allowing a before and after comparison of the map.

These results are preliminary and further research will continue, but the initial progress shows that there is potential in these methods. The early success provides the traction at NASA to continue funding of the research. The software written by the CMUSV students will function as a proof of concept, and provide the evidence necessary to continue funding research on these space flights. NASA engineers will take the source code and refine it to match the style and standards of NASA open source software before it is released to the general public.