Tip:
Highlight text to annotate it
X
Askins: Ares I-X used Windchill as our primary data sourcefor
storing data as well as working throughour
boards and reviews. At the beginning of the projectit
was extremely clunky and slow, mainly becauseone,
we were using it in a way thatit hadn't been chosen-
hadn't been designed to.We actually overfilled
the Windchill environment.Which even made the problem
worsebecause it slowed down, people got time-out errorsand they had trouble even accessing documents
at various times.We did learn, over the
course of the project thougha few real key valuable
tools. And Windchill is functioningfor
us now, in our data area, where one,we've
got segmented projects where we don't have themoverloaded
any more. We also have the number of usersand
the access limited to those that need itinstead
of having everybody access to everything.Because
that also puts a lot of overhead on itwhere
we were filling the buffers. We also used it for
our review process of our drawings.It
was very successful in using in that limited area,capturing
the reviews, and then putting outthe integrated drawings
with the- including those inputsfor the review. So all
that works very well. Sowe're using the tool now
in a more efficient mannerthat would have helped
us at the beginninghad we had some good training
and understanding of the tool. But we started
out using the toolbecause- as a directive.
So we had no trainingor anything up front. One
lesson that we really ran into at the end is
we allowed a lotof our IPTs, which are-
was our individual- ourour product teams to develop
some of theirtools and store them outside of the Windchill environment,which was the final tool that we needed everything
in.So at the end, there was a lot of clean-up activitythat we're still running through, to get it allback into the authoritative tool.It would have been a lot better for the programand easier access all the way alonghad
we all been working in one tool.And the one tool we happened
to use at the endand we'd been required
to, was the Windchill environment.The Windchill environment
works well in oneand we learned on that
is instead of having one major project, it's better
if you have a lot of little projects and link
them together.Because we actually were
the first to actuallyhit the threshold for the
Windchill environmentwhere it started slowing
us down.And that's mainly because we were way aheadof the mainline program. We were doing drawings.We were developing hardware. So our data-we were developing a lot more data a lot faster.So we sort of overfilled all the buffers, sothat was something we learned. That breaking it
apartand the mainline has done that. But up front,you know, everybody was trying to keep it inone project. We found the limits of that. So.Those are two key things, is not having everythingin one Windchill environment. Having it in andsegregated. But then linking them to the main.And then the other was one- keeping all the IPTsin one particular tool from the start of the programso that way you don't have a lot of close-outactivities. Our CM process we did manually in Windchill.So we really- huge amount of effort was put.Man hours, time, and everything. I really thinkthere should be a tool that makes that more automated.Because one, we were very much one-deepon
our team. And two, it was so manual. There's
got to be a way to dothere's got to be programs
and all.We just didn't have time to go look for them.So that CM function needs to be a automated processversus a manual process.Jeff Campbell: You know, another lesson that I thinkwas key was in 1-X, we were asked by Constellationto go ahead and try to come up witha
method to do paperless processing of the hardware.That
means not just "I typed it in on a computer,""I
printed it out, and handed it to someone else.""And
he went and hard-stamped it." That means"I
never used a piece of paper for the entire flow.""It was all electronic." Now, in Shuttlewe used some paperless processing on the ground
side.We have a product called Maximo. You know,that's for all of our facility-maintenance type items.And so, those folks have been using paperless
environmentsfor eight to ten years. In Shuttle, we had never done thaton
the flight side, so this was all new.And
it was a big challenge. So we had to go throughand make sure all of our business processescould accommodate this new tool. We said going inwe're going to use commercial off the shelf product.We're not going to modify it. Because any timeyou modify a commercial off the shelf productyou can do it, but there's going to be a cost. So we saidwe're
not going to change that at all.We're going to change our
business processes.So we modeled our business
processes to meet the tool.And we ended up launching
the vehicle successfullyat the end of October.
And we never usedone piece of paper to do
any work constructionson the floor at KSC. At
the other sites,you know, they had whatever
their Heritage systems usedin order to assemble the
various piecesbefore it got to KSC. But
once it got hereall of our processing was
done electronically.The engineers authored
the work instructions.They sent them out for
approvals.The technicians, quality personnel executed those
work instructionsand bought off all the
buys electronically.There was not one physical
piece of paperwe used for 1-X. And to
me, I mean,that was a huge success story. And the lesson isthat, you know, you can do thiseven with a complex system.It's
not like Shuttle, but you know,in terms of business processes,
it was no simpler reallythan Shuttle. We still
had to have USA, NASAand the contractors review
the paperwork.And we did that. We did
that successfully. So.And in my mind, when we
got down to reallythat last day, when we
had a lightning strike.You know, it was the night
before launch.We came in, and we had
to go throughand we had to basically do a littlea lot of retest in order to get flight rationale.We were able to do that, reallyBecause
of [INAUDIBLE]. Because we were ableable
to make real-time changes to the procedure.Everybody
could see what was worked.What the values were input,
and so we couldreally realize we've done
all the testingwe needed to do. We satisfied
all the requirementsthat we needed to meet
in order to get flight rationaleto the launch [INAUDIBLE]
team that day.So the lesson to me is
that you can usepaperless processing to
process flight hardware.And you can do it successfully.And the benefit- an overall benefitis that it's quicker to do changes to procedure.It's quicker to close out those procedures.And process control of the procedure you're workingis actually better, because the tools that are out there,various tools, allow you to basically put in limitsthat nobody can go beyond. So if you are processinga piece of hardware, and you violate a requirementit will not physically let you move outof that procedure.Mike
Stelzer: We tried to leverage Heritage tools justas
much as we can, and two of the ones that weborrowed
from Shuttle was our change management express toolour
action management express tool. And these worked outvery
well, especially in the later months of the missionwhen
we had a lot of changes, a lot of actionscoming
through. And what these management tools allowed usto
do is leverage the email system as a change came inor action came in, to get that distributed outto the folks who needed to review it.Get their comments in, get those comments addressed and incorporated.And be ready for the board, where we needed togo implement those changes. So those worked out
very wellleveraging those Heritage systems. So we've passed that
onto ourlong-term to look for those type of systems.Make sure they're available for the long-term program.One of the areas that we also leveraged was inone of our financial management tools. And this toolis based out of our contractor, and was really builtfor a quarterly reporting period. With Ares 1-Xthings were happening fast and furiousand
we really needed something that reported outat
the monthly, or even the weekly, level.One
thing we didn't realize in leveraging that, thoughis
using a reporting tool that's based on a quarterly,we
were trying to cut that a little closer.We
could make some decisions in between the quartersand
we realized that it really wasn't built for that.That
some of the estimates that were coming in beforewere
either low or high. So our recommendation,our
lessons learned to long-term, if you're usinga
tool like that, and you're using it differentlythan
how it was applied, in this case quarterlyversus
monthly or weekly, you have to look very closeto
make sure you're not making erroneous assumptionsin
there on the data in between the reporting periods.