DoD Has Space Management Problems Too
Space Acquisitions: DOD Continues to Face Challenges of Delayed Delivery of Critical Space Capabilities and Fragmented Leadership, GAO
“Many major Department of Defense (DOD) space programs GAO reviewed have experienced cost and schedule increases. For example, costs for the Advanced Extremely High Frequency satellite program grew 118 percent and its first satellite was launched more than 3.5 years late. Costs for the Space Based Infrared System grew nearly 300 percent and its scheduled launch was delayed roughly 9 years. Both programs are now in the production phase during which fewer technical problems tend to surface. Satellite ground systems have also been challenged by cost and schedule growth. In fact, ground system delays have been so lengthy that satellites sometimes spend years in orbit before key capabilities can be fully utilized.”
And people blatantly and publicly complained that the last and next space telescope are late and over-budget. Telescopes which have and will broaden human knowledge, more than possibly anything else, about the nature of our universe, rather than spy on other humans and help control and kill them. This is the first I have heard of more military waste. In comparision, humanity is really messed up.
The Federal Acquisition Regulation applies to them as well as NASA. I’ve never worked on a DoD contract, but I’ve been told their processes can be even more cumbersome than NASA’s. Do you know what a toilet seat costs when someone says, “anything going on a B1 bomber has to be able to survive a near miss by a nuclear bomb?” (Ok, I’m guessing about why they cost the Air Force that much…)
Having some knowledge of extremely high frequency radio, both reception and transmission, I wonder if the particular satellite given as an example was perhaps aggressively specified?
In fact would it be true that many missions are so ‘bleeding edge’ that the true difficulties of implementation aren’t understood at project initiation?
That can be a problem. It’s hard to say with DoD projects, since all the details aren’t public. DARPA has certainly done some things which were very far beyond the existing state of the art, and I’d be shocked if those projects always worked, let along worked on time and on budget.
But from the full GAO report, these all look like operational projects, and next-generation replacements for existing assets. I’d expect those projects to emphasize reliability, not push the state of the art. In addition to the examples Keith mentions, ground control for the next-generation GPS satellites is mentioned as having cost and schedule problems. I’m just not seeing it for things like that. The requirements could have been overly ambitious. But that should have been clear from the start based on experience with similar, past systems.
This is what caught my attention:
Not being close enough to the industry I wonder how often something like that happens. And there must be a dozen companies operating extensive ground systems – by my count, SES has 44 satellites operating!
Ground stations aren’t rocket science (sorry). It’s complex for sure but isn’t this a well understood activity by now?
The ground system can run into problems when you do something new or different. SES, for example, is shifting to all-electric propulsion (without a kick stage to take satellites from GTO to GEO.) That’s operationally different for the first few months after launch. I expect their ground system will have a few bugs and glitches to deal with. I believe they are also expanding into services like direct to user communications (e.g. satellite phones, ships and aircraft.) That means different sorts of data management.
Sometimes, “something new” can be subtle. Juno’s ground system had some problems in pre-launch testing, because the nature of the data from the instruments was subtly different. The system was designed to store and pass on the correct data volume at the correct rate. Storage space, memory, data rates between the DSN, LMA, JPL and the instrument teams were all fine. But the instruments were the sort which produced lots and lots of small data files (as in one, few kilobyte file every second, as opposed to a few megabytes per file once every ten minutes.) The overhead in handling a large number of small files wasn’t expected, and the ground system choked. That was all identified and fixed before launch, but it was a surprise.
New or unique systems also require some in-flight experience to get things right. Years for a planetary spacecraft or instrument isn’t shocking. (Although years usually means learning to get more out of the spacecraft than you originally though possible.) That’s why the short lifetime of some spacecraft (JWST is planned for five years and can’t last more than ten) concerns me. After a quarter century, people are still figuring out new tricks to improve the quality of HST observations.