a year and a half ago, I sent all of my people -- the support
contractors and the civil servants alike -- to risk management
training. It was part of an ongoing commitment to manage risk
effectively on my weather satellite program at the Goddard Space Flight
They broke into groups for a full day of training, and then they all
got together for a workshop to create a list of the risks we faced.
When I came into the workshop, I told them that they were free to
suggest any risk they wanted, but they needed to understand that our
senior management team was going to review all the submissions to
decide what was relevant.
"Your imaginations could go wild," I explained to them, "and you could
generate hundreds of risks -- that you get run over by a car, or other
things like that. We can't plan for situations like that. So don't
submit ridiculous things to us like, 'We're going to crush an
instrument,' or 'We're going to drop the spacecraft.' Those just aren't
At the end of the day, we collected all the risks they came up with,
and we entered the credible ones into our system for tracking. We
reviewed some of these risks every other week and revisited the entire
list periodically. We were doing what we could to manage risk on the
program -- or so we thought.
|A 3,000 pound spacecraft dropping three
feet onto a concrete floor gets damaged.
You're Not Going to Believe What
On September 5, 2003, my wife and I left to go on vacation. We planned
to spend two weeks wandering around New York State seeing all the
sights. When we left the house, I turned off my cell phone, but kept my
pager on -- in case anyone needed to get hold of me. We had a wonderful
weekend. Then, bright and early on Monday morning, my pager went off.
It was the Project Manager for one of our spacecraft. She had been
trying to reach me on my cell phone since Saturday to tell me that the
day after I left, Lockheed-Martin had dropped one of my spacecraft.
You can go through your whole career and never have someone drop one of
your spacecraft. I think that would have been nice. So, one of the
first things I did when I got back, was to inquire whether I could
retire retroactively to Friday, so it wouldn't have been on my watch.
They just laughed that off.
Then we got to work. Almost immediately, four investigation teams were
formed -- two by Lockheed-Martin and two by NASA. Each was tasked to
investigate a different aspect of the accident. These aspects included
not only finding out what happened, but also looking for systemic
problems in the program, determining next steps, and assessing
What Went Wrong?
The "what happened" investigation didn't take long to report its
findings. To begin with, the procedure called for eleven people to be
present for this operation. There were only six there. The
Lockheed-Martin people had decided some time earlier that three of them
weren't really needed -- but they had never redlined the procedure and
notified us. The other three hadn't been scheduled. The safety guy
wasn't even notified, even though he was listed in the procedure.
The operation was scheduled to begin at 6:00 a.m. They also should have
had a NASA QA guy there, but when they called him, he said he'd be in
later and to proceed without him. When the contractor's QA person
arrived at about 6:30 a.m., they were on step six of the procedure, and
he said, "Oh, you're on the sixth step. Let me sign off on the first
five." And he stamped them, without bothering to look at anything.
One of the procedure steps involved inspecting the cart to make sure it
was ready to take the spacecraft. The test conductor said he used the
cart a week ago; so what could have happened since then? He didn't
Then one of the technicians went over to the cart as they were lowering
down the spacecraft. He told them something looked different, but the
test conductor didn't go over to look. He just said to go ahead. Turns
out, there was a ring of bolts missing. That's what looked different.
|We thought we were doing what we could to
manage risk on the program.
There were many steps bypassed that day, any one of which would have
caught and avoided the problem. They ignored them all. They went on.
They mounted the spacecraft. Then they went to turn it over on their
dolly, and it hit the ground. A 3,000-pound spacecraft dropping three
feet onto a concrete floor gets damaged. How damaged was a bit more
complicated, but estimates ran up to $200 million.
|Most spacecraft are safe on their dolly:
workers separate upper equipment module from its dolly.
After the Mishap Investigation Board (MIB) draft report came out, the
test conductor and two other people got fired. It was Lockheed-Martin's
response to show that they wouldn't tolerate this kind of activity. The
way I saw it, the people who got fired weren't necessarily the people
who should have been blamed, because they weren't the root cause of the
accident. I felt the blame really should have gone higher in the
organization. The Project Manager was replaced six months later.
There were several MIB conclusions with which I took issue. For
instance, they put some of the blame on the government, because we
didn't have our own QA person there at the time of the occurrence. I
believe that I should have reviewed all of the procedures and to have
made certain that things were in place for the contractor to do the
work properly and safely -- safely for the people, and safely for our
They suggested that we needed to have a civil servant in residence at a
plant for every project like this. But I don't think it matters what
badge someone wears. He or she just needs to have the right dedication,
the right training, the right experience, the right everything. Being
civil service doesn't mean a damn thing. I have actually used civil
servant leads and contractor leads at one time or another in the past.
Either will work as long as you have the right person in the right
But I was told by my management, "You will implement everything that is
in the report." No discussion, no exceptions.
Around that same time, I got my copy of the Columbia Accident
Investigation Board (CAIB) report. After reading it, I called my deputy
center director. I said that the CAIB Report tells me not to blindly do
things that I think are stupid. So, I said we needed to talk about the
MIB report. He started to laugh. Then he said that he would have to
think about that one.
So, we had a little standoff. Since that time, I have spoken with the
chairman of our investigation board. I found out that the MIB team
didn't unanimously agree to the things that I had problems with. The
next time they meet in Washington, as a complete team, I'm going to get
to talk to them.
|Another safe transition, the Apollo Spacecraft
017 Command Module is lowered onto a dolly.
An Incredible Risk Repeated
A risk (dropping a spacecraft) that I had summarily dismissed as "not
credible" at our risk management workshop actually has real-world
precedence -- both before and after our own event.
In mid-2000, another contractor, let's call them Contractor-B, dropped
a spacecraft. You didn't hear too much about that, because it wasn't a
government contract; it was a commercial contract. They dropped it
because of bolts that were missing in the dolly. (Sound familiar?) We
knew they dropped it, but the details never came out.
The same Contractor-B dropped another spacecraft in the middle of
December 2003. That made it into the Space News without much detail.
They had just run a thermal vacuum test on the spacecraft in Seattle,
Washington, and then dropped it while putting it back into the shipping
container. Someone was hurt in that accident.
None of these are simple cases where a team missed one step and so the
accident happened. It's always a combination of skipped steps or
miscommunications or dangerous assumptions. So, how do we mitigate this
sort of risk?
First, we need to properly identify the risk. In our case and the two I
sited above, the real risk wasn't necessarily "dropping the
spacecraft," even though that was the end result. The risk in our case
would more accurately be called "complacency."
We had a long-term project with our contractor. Their attitude was that
a spacecraft lift was not a risky thing. After years of doing this
work, they saw it as very low-risk. But, in truth, it's always a
hazardous operation. It should never be considered low-risk. It always
requires the full attention they gave it the first time they did it.
I've come to realize that, no matter how long you work in this
business, new experiences will keep coming along. Each one broadens
your horizon and helps you do better.
- Safety requires strict adherence to
- However, adherence to procedures
in repeated operations also requires the careful attitude typical of
To what extent is adherence to procedures -- coupled with the right
attitude, but unsupported by the proper experienced-based judgment --
sufficient to prevent known risks, but insufficient in preventing the
Search by lesson to find more on:
Marty Davis is the Program
Manager of the Geostationary Operational Environmental Satellite (GOES)
at the NASA Goddard Space Flight Center (GSFC) in Greenbelt, Maryland.
He is responsible for the design, assembly, integration, and test of
the GOES flight and ground support hardware and for the launch activity
and on-orbit checkout of the spacecraft.