[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

X-34 Program Cancellation

  X-34 Is Dead -- Long Live X-34!  Analysis & Recommendations

The Orbital Sciences Corporation/Rockwell/NASA X-34 air-launched partially
reusable lightsat launcher project is dead.  The short version: The project
had ongoing weight-growth, cost-growth, and schedule problems, OSC wanted
to radically downsize the vehicle, Rockwell quit, and NASA pulled the plug.
RIP X-34, and good riddance.  (We expected this - we were pointing out
problems with the overall project and refusing to suppport it over a year
ago.  See SAU's #47 through #50, November '94 through April '95.)

So it will no doubt come as a surprise that SAS considers the X-34 project
to date a rousing success, and has hopes for it doing a lot of good for US
commercial reusable launch capability in the relatively near future.

Allow us to explain this apparent contradiction.

X-34 as conceived by OSC/Rockwell/NASA had two major problems that combined
to kill it.  

One was obvious from the start: Too many agendas.  Just to list the overt
conflicts (never mind the covert agendas that various sponsors appeared to
be pursuing), X-34 was going to experiment with new and untried NASA-
contractor cooperative management techniques, at the same time it was going
to "prove" NASA could routinely work effectively with industry in fast low-
cost X-projects.  X-34 was also going to produce a cheap near-term
hypersonic flight testbed for reusable rocket components, and at the same
time be a low-cost commercial lightsat launcher.  

Developing new cooperative management techniques was a good thing to do.
In fact, that's one of the reasons we're calling X-34 a success so far; not
only did the project uncover and initiate reform of a number of internal
problems at NASA, but the new setup allowed X-34 to fail for only $8
million, rather than the several hundreds of millions such a project would
have eaten under traditional arrangements.  More on that in a bit.  

Operational and "X" objectives, now, just don't mix well, period.  X-
vehicles should not have set payload requirements, as that tends to get in
the way of building something to fly as cheap, high, fast, and soon as
possible, something that'll return the data needed to minimize development
risk for operational payload-carrying followons.  The payload requirement
had much to do with X-34's troubles.  More on this later too.

The other problem was one a lot of us thought likely (but of course not
certain) given OSC and Rockwell's combined track record for cost growth,
weight growth, and schedule slippage on complex aerospace vehicle projects:
Bad management.

Actually, "bad management" is an exaggeration, quite unfair to the people
involved - the problem is more accurately described as the lack of
outstandingly good management of a small highly-skilled highly-integrated
development team.  We'll more accurately call what OSC and Rockwell brought
to bear ordinary government-contract aerospace management - not bad, but
not good enough to meet all the agendas, on time, on weight, and on cost.  

Here's why we said that failing for only $8 million from NASA is a success
for X-34: Historically, such projects when totally government-financed will
pour good money after bad for a long time, despite cost growth, weight
growth, and schedule problems.  The fact that NASA's share was fixed at $70m
and any overruns came straight out of OSC and Rockwell's pockets
concentrated their minds wonderfully when the project started going bad.  

It became obvious to Rockwell and OSC that they were going to lose money,
they dropped out, NASA learned a lot in the process, and still has $62
million of the original $70 million to try and do something useful with. 
That sounds like a success to us.  

 - OK, What To Do With $62 Million?
Dan Goldin asked us a year ago what NASA could do to encourage small
aerospace startups without crushing them by sheer weight of government
attention.  We didn't have a good answer for him at the time.

Now, we just might.  Bear with us for a bit while we lay out the logic.
Many of the pieces of the puzzle are already present in the story of X-34
Mark I's demise...  

History tells us that rapid cheap development of effective advanced
aircraft/aerospace vehicles tends to be done by small, highly-integrated
skillfully-led development teams.  

Absent wartime urgency, the post-WW II development approach of mass-assault
divide-and-conquer engineering tends to produce expensive protracted
bureaucratic bogdowns, the vehicles resulting (if any) ending up as complex,
fragile, difficult-to-operate kluges.  

We are conspicuously short of effective small development teams lately.  
This is because we haven't been doing much to foster such teams for a long
time now - neither encouraging the general approach, nor building many
specific projects of the sort such teams can form around and learn on.  

We're obviously still capable of doing this sort of thing; two recent
examples are the DC-X and Clementine developments.  Note that both teams
despite their successes have been in danger of being broken up and scattered
to the four winds - current management theory about workers all being
standard interchangeable parts is very destructive in this regard, as is the
overall lack of small/medium development projects.  
The US very much needs such development teams in the high-performance
aerospace vehicle field, in order to affordably exploit recent technology
advances and bring access costs down radically enough to make space an
explosively expanding US commercial sector.  The budget climate for the
forseeable future rules out any Cold War-style government-funded mass
assault on the access cost problem.  Probably just as well...  
It is very difficult for such a small development team to both advance the
state of the art, and at the same time produce an operational (reliable
payload-carrier) vehicle.  Imposing a rigid payload requirement while there
is still limited hands-on experience with the new technology tends to cause
the project to bloat, cost and weight-wise.  Large unknowns require large
engineering margins, if the result must be reliably operational.  
Such new ground is best broken by building an X-vehicle, one designed to
explore the flight regime quickly and cheaply, with relatively high risk of
initial performance shortfalls and operational difficulties an acceptable
tradeoff for getting results fast on a tight budget.

Once a small development team has explored the new flight regime with an 
X-vehicle, they're ready to move on and produce an efficient operational
followon, quickly, cheaply, and with relatively low development risk.  

But such an X-vehicle flies no operational missions, carries no payloads,
makes no money for anyone.  Profits come only after the second 2-3 year
design-build-fly iteration.  There's no payoff for at least 5-6 high-risk
years from cold start, so commercial funding isn't available for projects
that require advancing the state of the art.  And X-34 Mark I has again
demonstrated the unwisdom of trying to compress the process into one design

It is more than a little ironic that successful development of affordable
single-stage reusable space vehicles requires going back to the traditional
multi-stage advanced development project...

 - So What Should NASA Be Doing?
SAS believes it is a legitimate role for NASA to sponsor relatively small 
X-projects, with the specific purpose of fostering such small development
teams, allowing them to prove their management/technical competence by going
through the initial design iteration of flying an advanced X-vehicle, as a
precursor to a potential low-cost reusable operational spacecraft.  

SAS believes that a design team that has successfully done this will then
very likely be able to obtain commercial and/or customer funding for an
operational followon, given that they will have established relatively high
confidence that they will be able to complete the second (money-making)
design-build-fly iteration on time and on budget.  

 - How can NASA best do this, given $62 million in the kitty?  
We don't have a sure-fire formula, but our advisors have come up with some
sensible suggestions...  
Go for one project for now.  $62 million is likely barely enough for an
extremely lean-and-mean outfit to get one X-vehicle flying.  (One useful
possibility would be to accept a vehicle bid in tandem with a bid to build
an upgraded version of a proven engine, using some of the small-engine money
we hear will be materializing at NASA.  This allows a more ambitious vehicle
for the available money, while providing a fallback position of initial
flight with the old proven engine version if the upgrade is delayed.)

Do this as a small-business setaside.  Nothing against the existing outfits,
but there's already a project to foster one or more skilled reusable rocket
design teams in the existing large aerospace corporations; it's called X-33. 
This one's for the startups.  (The big guys, frankly, should be using more
of their own money to build and maintain strong design teams.)

Pick the winning bidder primarily on the strength of the team they've lined
up, on the balance between management, design, fabrication, and operations
skills, next on the strength of their plans to market their technology if
their X-vehicle succeeds, with their financial backing a distant third. 
We've seen there'll be little commercial money for such ventures at startup.

Insist on the winning team being committed to success.  They may not have
much for assets and financing, but what they have should be at risk.  Make
the CEO mortgage his house; set up the payment schedule so that if the team
is reasonably close to on-time and on-budget, he can pay it off.  Make team
members contract for the length of the project; no bailing out for a better
offer or from disgruntlement halfway through.  
Make the contract such that if the winning team does a decent job, they end
up with enough for a month's screaming hedonistic vacation (they'll need it)
followed by a year or two working up their operational followon design and
looking for funding and customers.  If they screw up, at most cover the
remaining debt after they sell off their assets.  
Monitor their technical progress and expenses closely, but supportively. 
This is critical!  We've heard from numerous entrepreneurs about bad
experiences working for the government.  Some of this of course is simple
aversion to anyone at all looking over their shoulders, to doing any
paperwork at all.  But much of it stems from being burned by letter-of-the-
law beancounters/procurement-paperwork enforcers who neither knew nor cared
what the project was trying to accomplish.  The technical and financial
oversight people should be knowledgable about the project and supportive of
its goals.  They should be there to spot problems early and work with the
development team to solve them, not to shout "gotcha!" 
Allow the project to fail.  There should be some slack in the schedule and
finance, as long as the reasons for using the slack are clearly understood
and under control, but if things get out of hand, don't pour good money
after bad.  Shut it down and try again elsewhere.  

Success or failure, follow this up with more small projects as funding can
be found.  Such projects are inherently risky; the way to ensure the big
payoff is to make multiple small bets.  

This advice is aimed at partially/fully reusable smallsat launcher
development in this particular instance, but we think it's more widely
applicable.  And that's all we have to say for now.  Until #62...

 Space Access Society      "Reach low orbit and you're halfway to anywhere 
 4855 E Warner Rd #24-150               in the Solar System." 
 Phoenix AZ 85044                               - Robert A. Heinlein 
 602 431-9283 voice/fax 
 www.space-access.org                     "You can't get there from here." 
 space.access@space-access.org                          - Anonymous 

 - Permission granted to redistribute the full and unaltered text of this -
 - piece, including the copyright and this notice.  All other rights      -
 - reserved.  In other words, crossposting, emailing, or printing this    -
 - whole and passing it on to interested parties is strongly encouraged.  -

                      (from www.rlv.msfc.nasa.gov)
+                                                                             +
+  Weave a circle 'round him thrice, and close your eyes with holy dread...   +
+                                                                             +