Political Commentary, Space

The Democratizing Effect Of Social Networking

18 Comments

So I tweeted thusly a few minutes ago:

Rand_Simberg Rand Simberg
Wouldn’t know it by her committee actions. @spacelawyer “I’m a lawyer and I do want to go to space.” – @RepDonnaEdwards

And lo, a few minutes later, she or whoever actually runs her Twitter account replied:

repdonnaedwards Rep Donna F Edwards
@Rand_Simberg pls don’t misunderstand tough questions & vigorous oversight. My goal: build a strong, competitive, innovative space program.
Jun 06, 11:39 AM via Twitter for BlackBerry®

FWIW. I’m tempted to tweet back that SLS will never get her (or likely anyone) into space. Props for responding, though.

18 thoughts on “The Democratizing Effect Of Social Networking”

  2. I’m still trying to fathom a paragraph in this article about the arrival of SLS flight computers.

    Dane Richardson, manager for the Boeing SLS Avionics and Software Team, said “These are the most capable flight computers ever developed for human spaceflight. They have the highest processing capability available in a flight computer and triple modular redundant processors. The technology is proven from years of satellite applications, and it’s reliable enough to take SLS beyond Earth’s orbit.”

    If they’ve been used for years in satellites, they weren’t really developed for human spaceflight, were they? And if they’ve been used for years, they can’t quite be state-of-the-art CPUs.

    That goes along with this quote:

    “The final SLS flight computer that will run the flight software will have the highest processing capability available in a flight avionics computer.,” NASA said. “It is being developed by upgrading existing systems used in Global Positioning System and communication satellites.”

    And how much processing power does it take to control the SRB’s? An ignition signal and two steering outputs is pretty much all you’ve got to work with. They just burn. The RS-25E’s and J2’s were happily controlled with 1970’s computer power, so I don’t think that’s quite it, either.

    1. These will play Missile Command at the same time. Try that with 70s vintage computers…

    3. And if they’ve been used for years, they can’t quite be state-of-the-art CPUs.

      In his defense, he probably meant earlier versions of the lineage they are using. Many such things are steadily upgraded over the years. So it is possible for a CPU that’s never been in space to be a more advanced version of one that has.

      The story you linked to is remarkably empty on details with the only real information provided by that Dane Richardson quote.

      I don’t know what the point of a more advanced avionics system is in this case, but I imagine it has a lot better data logging than older systems. So for example, one might have a mesh of hundreds of sensors reporting on vibration, displacement/bending/load, and temperature over the entire outside surface of the motor in millisecond (perhaps even in microsecond) increments and store that information in a small box that you could carry in your pocket. I don’t really know the potential capabilities of such avionics, but a key one is simply providing a detailed log of the vehicle in motion. For rockets, a lot of things happen fast, and being able to dissect problems and accidents into very fine detail would be extremely useful.

      1. I believe Boeings rad hardened PC is based on the PowerPC 750, as are offerings from AiTech and BAE systems. AiTech’s S950 specs are probably pretty comparable. Their new S960 should be out, said to be plug-in compatible with the S950, which implies that a newer version of PowerPC architecture has passed muster. I suspect NASA is using the Boeing equivalent upgrade.

    5. I wonder whether the arrival of the SLS flight computers is being treated as a milestone because the delivery of the first Apollo GNC from MIT’s Stark-Draper labs was a milestone, as was the development of flight computer capable of flying the Space Shuttle. Computing has made notable strides since the 60’s and 70’s, which is probably why SpaceX, Armadillo, and Blue Origin didn’t have a press release about buying three computers long before they’d built flyable hardware. Everyone would’ve laughed at them.

      1. George,

        If you read between the lines — ” reliable enough to take SLS beyond Earth’s orbit” — it’s obvious they’re talking about rad-hard computers.

        Rad-hard chips are so far behind the state of the art, you can’t see them in the rear-view mirror.

        Rad-hard components are unnecessary in most cases. There are other ways of ensuring reliability, such as spot shielding and watchdog systems that automatically reboot the computer when a radiation fault is detected.

        Even when rad-hard components are required, there’s nothing magical about procuring them. It just requires a lot of money, a lot of patience (there’s a long lead time on orders), and the willingness to live within the hardware limitations.

        The main reason rad-hard chips lag so far behind consumer electronics, in both performance and cost, is the fact that they’re produced in minuscule numbers.

        There’s a lesson there for launch vehicle designers, which goes beyond electronics.

        1. Radiation hardened CPUs lag way behind the state of the art terrestrial CPUs but massive computational power isn’t needed for most systems. There just isn’t a lot of calculations being performed. Why pay for something you don’t really need?

          Back in 1982, the Air Force launched the first military communications satellite with an on-board CPU, the DSCS-III. The CPU was radiation harded in part because of the space environment but also to better withstand a space nuclear detonation. The CPU was a 16-bit chip derived from the DEC PDP-11. There were two of the chips (one backup) and two banks of RAM, each with 2K. That was sufficient to maintain attitude control of the satellite, perform routine automatic functions like momentum wheel unload thruster firings and control the beam forming networks on the uplink and downlink multibeam antennas.

          Later satellites contain more powerful CPUs and greater amounts of RAM but not massively more than the old PDP. It just isn’t needed. What is it going to do, wait faster for something to happen?

          1. Radiation hardened CPUs lag way behind the state of the art terrestrial CPUs but massive computational power isn’t needed for most systems. There just isn’t a lot of calculations being performed. Why pay for something you don’t really need?

            Preprocessing large amounts of data “out there” can be very useful because it reduces the amount that needs to be communicated back to Earth. That will be important for low-cost probes that can’t afford to rent the Deep Space Network 24/7.

        2. As I understand it, there’s also a technical reason the rad-hard chips are so limited: a big part of why they’re rad-hard is that all the individual transistors are made big enough that a single-partical hit won’t flip its state. This means they’re really big by modern standards, which means they take a lot of space and use a lot of power because they’re forever stuck on mid-80s process technology.

      2. If they went with COTS there are several possibilities. Just as a wild guess, perhaps an AiTech S960, which should be coming out about now. Here are the previous S950 specs, which is a 733Mhz PowerPC with 128MB of RAM and 64MB of flash memory. But since the computers were delivered by Boeing, they’re probably not COTS.

  3. “Most capable” is an interesting phrase.

    Consider this NASA drawing comparing the Space Launch System to other rockets, including the Masten XA 1.0 and XCOR Lynx.

    http://www.citizensinspace.org/wp-content/uploads/2012/06/size_comparison.jpg

    SLS is the most capable because it’s the biggest, right?

    Now, take a look at this drawing, and tell me which airplane carried more cargo.

    http://www.citizensinspace.org/wp-content/uploads/2012/06/WWII-comparison.jpg

    1. ^_^ Love that WW-II comparison.

      But the SLS is bigger, and it has electrolytes!

      They go with liquid hydrogen instead of RP-1 and gee, their rockets are so much bigger! Given that the Falcon 9H can put 60 tons to LEO, you could stick two of them side-by-side with a common upper stage and equal the payload of the Apollo and the giant 1301 SLS, and probably do it for less cost than the purchase price of just the four throw-away SSME derived engines on the first stage of the SLS.

Comments are closed.