In 2007, a small group of people began an intentional, collaborative experiment in open, transparent, and direct communication about your space program. Our goal was to enable your direct participation in exploring and contributing to NASA’s mission.

Many of us have since begun new adventures. This site will remain as an archive of the accomplishments of the openNASA experiment.

NASA’s Liquid Galaxy: An Overview

By Stuart Engelhardt, 26 September 2011
 Stuart Engelhardt

[youtube http://www.youtube.com/watch?v=qbin9WUHrDg]

A few months back, Open.Gov Team Lead, Nick Skytland approached me with an idea to implement Google’s Liquid Galaxy project at NASA.  I was immediately intrigued by the idea, and set forth researching what would be involved.  Nick and I collaborated with others at Johnson Space Center (JSC) to come up with a way to implement our own Liquid Galaxy.  We also collaborated with people from other NASA centers to come up with ideas to extend the project’s functionality, and to apply the same hardware to other activities.  In the end, we came up with an exciting eight-display Liquid Galaxy “immersive cave” and a number of ideas to consider for the future.

Early on, we sought to participate with the JSC Innovation Team, and were successful in being selected as one of 20 innovation projects they wanted to see implemented.  This gave us access to other innovators in the center, to collaborate with, and better utilize available resources.  Thankfully, we were able to get an early start on hardware acquisition, which proved to be the most critical scheduling component for almost all of the 20 innovation projects.  Over the course of six months or so, we were able to participate in a variety of innovation activities around JSC, including a Maker Camp, and several innovation project progress presentations - which gave us great insight into the brilliant minds and ideas others brought to the 2011 JSC innovation effort.

During our project’s implementation, we collaborated with a variety of people on different aspects of the task.  During the two day NASA Forward - Maker Camp in August, people from Space Life Sciences and Engineering Directorates joined in with designing an 80/20-based rack system to mount the hardware vertically.  This particular solution gave us a configuration that will work for either a HyperWall or Immersive Cave configuration of the Liquid Galaxy.  We also worked with people from Mission Operations Directorate and the new Commercial Crew & Cargo Program Office (C3PO) to overcome hurdles we encountered with system configuration and optimization.

To get an idea of what we could do beyond Liquid Galaxy, we talked with a variety of teams involved in graphical simulation projects at different NASA centers around the country.  Our first conversation was with the group responsible for NASA’s World Wind project at Ames Research Center in Moffett Field, CA, who we had met earlier this year at the NASA Open Source Summit.  We also talked with the VR Lab Team at Johnson Space Center in Houston, TX, who created DOUG - a 3D viewing tool used in training for space walks as well as shuttle and station robotic arm operations.  Finally, we have spent the most time discussing opportunities with the team responsible for GMAT, the General Mission Analysis Tool, from Goddard Space Flight Center in Greenbelt, MD.

Discussions with these groups looked at two different possibilities for machine clustered HyperWall and Immersive Cave implementations.  The first is similar to Google Earth’s implementation, where you run the application on every machine, and have one configured as a controller that sends data to the rest, and the others receive the controller’s state data (position and attitude) and then draw their locations based on offset values applied to the camera / screen position.  This requires the software on every machine, and in the case of Google Earth, each machine is requesting all of its own data from Google over the network - in our case, eight times the bandwidth usage.

The second option is an innovative solution recently mentioned on Google’s Liquid Galaxy page, by one of their three GSoC (Google Summer of Code) Interns: ClusterGL.  ClusterGL is basically taking OpenGL, an open source graphic library alternative to DirectX, and providing its graphics API calls over the network so that other machines can draw the graphics themselves, without needing application on their machine.  You can run one copy of the program on one machine, and setup all the rest to just draw additional screens of data to make the field of view larger.  So far this technique shows the most promise, because you don’t need to adapt an application to talk between multiple copies of itself on the network.

So far our project has been met with great interest and enthusiasm on site at Johnson Space Center.  People just can’t stop playing with the Liquid Galaxy, flying around the Earth, Moon, and Mars, looking at all of the graphic visualizations of the world in an Immersive Cave environment.  If you are ever at JSC, come over to Building 29 and try it out.  We look forward to working with NASA and external parties interested in collaborating on clustered graphical simulation solutions in the future.

By Stuart Engelhardt