Congressional Hearing on Safety

I watched it...it was an Ares Luv Fest. Talk about "wiring the message". :)

Despite all of the talk about increasing NASA's budget, this subcommittee has a pretty bleak track record of convincing the leadership (when either parties were in power)of doing same.
Let's hope the President backs the budget boost and directs the Democratic leaders to do the same.

For those interested, the URL for the hearing follows:

http://science.edgeboss.net/wmedia/science/scitech09/120209.wvx

Rohrabacher made a fool of himself at times especially when he was shilling for Boeing.

The only way to guarantee perfect safety is to never fly anything.

There is some insight I can offer on this on-going debate about the safety metric in human space flight, specifically in its latest rendition regarding Ares I, as well as in terms of COTS-crew initiatives, or in light of the Orlando Sentinel’s story. Indeed there is quite a “story” to be told here as it mixes up the basic elements that make for a good story, people, money and power.

For one, yesterday’s House hearing on Human Space Flight Safety showed par-excellence the continuous distraction, side-topics and “on the other hand” type of wandering debate that any discussion on a crew safety metric always seems to produce. I’m surprised the committee members did not end the discussion by asking that future program managers and analysts that testify have just one arm.

It’s a question in itself worth asking - why something so seemingly simple as losing or not losing a crew, with no grey area in-between, gets so bogged down in such a public forum as to be nearly un-intelligible to committee members and lay-persons alike?

First, the real requirement that the Constellation program has that is referred to often when speaking of an “order of magnitude improvement over Shuttle”, as regards safety of crew, is the end-to-end mission requirement. That is, the requirement is 1 in 1,000 probability of loss of crew, from launch through landing. It applies specifically to a mission to the International Space Station. The term “order of magnitude” is a notional term, sometimes used in referring to this whole metric, but just as often used when talking only about the ride-up, the launch itself, which is just part of the “mission” (which is the ride-up plus everything else, including the return landing). To put this in some perspective think of the Shuttle as having demonstrated on the ride up a vehicle (and crew) loss rate of about 1 in 100 and then you have the 1 in 1,000. An “order of magnitude” adds a zero.

But let’s come back to the Constellation requirement of 1 in 1,000, for the entire mission, which means not just Ares 1, but also Orion on–orbit and Orion’s return landing. This value made it into the fight, and I’ll call it a fight because as events have shown it causes a stir even to this day, as a result of the passions of some members of the community who don’t want to hear anymore excuses. Yes, I would say I’m one. It’s a simple yet tangible metric, connecting well to many other goals that are at the root cause of access to space being routine and accessible to many versus being a strictly governmental affair for the deep pocketed. And that’s where the heat comes into the debate.

When Columbia was lost, then President Bush stated a Vision, yet the Vision never actually emphasized safety on that day in January 14, 2004. If anything it emphasized risk taking and continuity. “The legacy of Columbia must carry on”. This is ironic given the Vision was a direct result of the loss of Columbia. The Vision never said the next system would be much safer. This tidbit developed later, within the Constellation programs “Needs, Goals and Objectives”. And herein began the battle. (You’ll need to Google “Constellations Needs, Goals, Objectives”)

When the meaning of the Vision had to start being defined more tangibly, when “how” was being defined, many of the team members had participated in previous NASA studies or Shuttle replacement efforts. From the late 90’s the “order of magnitude” improvement in various metrics had been placed into programs from X-vehicles to Space Launch Initiatives ad nausea. A surviving, very ambitious improvement goal that made it into Constellation was crew safety.

And here is where the plot thickens. The community of assorted advocacies saw an opportunity to preserve some ambition in the goals, in the “ilities” of the next program, and they took it. With Columbia fresh in the collective memory such an ambitious crew safety goal was inserted rather easily in the proper documents, in all the right places. To boot, energies were organized around this strategic goal while giving up on others (like reliability) knowing it would be relatively easily married into the programs documentation and party line, and that this safety goal would actually drag back in other goals that were impossible to push at the time. Pushing for improved cost goals or requirements, for example, has failed left and right. Knowing that the Constellation program was not a done deal, the advocates of the ambitious 1 in 1,000 loss of crew probability on an end-to-end mission to the Station, knew that the requirement would be embraced quickly by a program still trying to sell itself. The guess was correct, and so the requirement got in. The “sell” was on, and safety became part of that. The advocates of the safety requirement also guessed (I’d say correctly as the hearing showed) that this strategic requirement could be much more easily inserted into the program than it could be discarded. The ambitious crew safety requirement would be sticky.

Now a crew safety metric can be worked in many ways. A Launch Abort System is one way to take a system and improve on crew safety. This would be the “it blew up but we got the crew outta there” approach to safety. But that only addresses the ride up, the first 100 miles as the term was used in the hearing. This tack was predictable, so the crew safety metric inserted was end-to-end, guessing that the program would not want to improve any critical hardware reliability, nor perform any more testing and re-design in test-fail-fix cycles than previous programs (these improvement as requirements could not be sold head-on). It could be counted on that the knee-jerk response of the program would be to let the abort system forgive all sins. Hence the inserted requirement had to be end-to-end, for the whole mission to the Station.

In practice this is what the program is attempting to do, yet the end-to-end requirement is still alive and kicking. Witness the way in which the Constellation program is trying to turn attention to the ride up, to the first 100 miles, to the climb. This was predictable. Hence the reason the “Ares I” gets mentioned in the hearing, as that’s the part that might fail on the way up, much less so Orion. An end-to-end view would say the story is not over at the end of the ride up. And herein lies the real requirement, the 1 in 1,000 loss of crew probability, end-to-end, for the entire mission to the station, which of course includes Orion, gets dominated by Orion and is calculated all the way to touch-down and crew retrieval.

Now there are better ways to go about setting metrics for improving and making more accessible the means of access to space (the technology base, the knowledge base, the business model, etc). Yet lacking an embrace of such broader goals, a metric pushing improvement on the end-to-end mission to the Station loss of crew probability to levels far higher than Shuttle is a way of dragging along, by hook or by crook, the other improvement metrics. For example, the reliability of critical hardware must be improved to achieve better crew safety end-to-end. Chances are that if the reliability of critical hardware has been improved then the overall quality of systems is better. Chances are this will reduce failures on the ground during processing. Hardware that is very good at performing in flight, predictably and at high reliability, likely will need little attention during its weeks after delivery from the contractor’s manufacturing plant. So crew safety breeds ground processing reliability. Chances are this increases flight rate at a given cost. So affordability on a recurring basis is enhanced. This too is a natural outcrop, as the test-fail-fix cycles meant spending more up-front, which links to spending less on a recurring basis later. All this leads to sustainability – a productive system characterized by an improved cost picture (fixed and variable), an improved reliability in processing (fewer brokes) and better flight rate at a given cost (productivity).

This is the really, really complicated way of saying that so long as human space flight is characterized as something filled with both uncertainty and risk then it will never be affordable (as we try to inspect in what was not tested in, nor built in) or routine (as the designed-in reliability is again lacking) or productive (as flight rate suffers at the mercy of costs planned and un-planned).

This leads to COTS-crew. Knowledge has to spread. A mature industry is also, by definition, one where the product’s knowledge has spread far and wide. NASA has very few mechanisms by which to spread knowledge in the world of ITAR, company proprietary, winner-take-all-competitions. On the contrary, the NASA knowledge model being contractor based once awarded, knowledge gets scurried away, locked up, and everyone else gets told to go home.

Regardless of the outcome of COTS-crew, this broad goal of spreading the knowledge, meaning a certain redundancy in efforts, is absolutely required to mature the industry. Maturity means knowledge spreading among many parties. The ambition of human space flight and its safety requirement will force in-direct improvements across the board in technology, quality, and business processes that a cargo initiative would never drive. These improvements one-step removed from safety (cost, reliability, quality of hardware, people processes and business processes, flight rate, productivity, infrastructure) are dragged along by safety. It’s a fight that has to go on to have NASA fund competing and distinctively modeled proposals under any budget environment. COTS-crew may fail in the near term yet it should never be abandoned in its basic reason for existence. And should COTS-crew initiative succeed then the road beyond Low-Earth-Orbit will be that much more easily explored, built up with cargo, and funded.

In a world of limited budgets power in a program means thinking that doing means a set of limited, near term goals. I’m taking the crew up, not you. Regardless of budgets, an Exploration program should never have been formulated absent the realization that knowledge has to spread, absent planning on the funding for COTS-crew out of foreseen budgets. Yet even when seeing far more generous budgets years ago this was not the case. The “winner take all” notion has dominated both Exploration at the high level, and Constellation below. This has to stop. Broader NASA human space flight goals have to be accepted and acted upon. So here we are. Debating endlessly in purposeful confusion something so simple as safety. It’s testament to the lack of advancement in getting to Low-Earth-Orbit that we are having such a debate. It’s testament to lack of past investment. The question becomes – shall we define safety and related goals and then see what we can afford, with reasonable certainty, lots of margin and with an awareness of the way budget promises change too easily. Or shall we hide behind “it’s complicated”, “at the proper funding level – then maybe”, and so on?

If we are still having this debate as the years go by it says something – what are we doing going beyond Low-Earth-Orbit when even putting a safety metric, or a cost, non-recurring or recurring, or a schedule, to getting to Low-Earth-Orbit is like nailing Jello to the wall?

Some things come first.

Excellent post! Thanks.