"I left the room and I thought about two things," Pellerin says. "All the trauma that was around this, and it wasn't going away. I mean the international partners are mad at us, NASA looks bad to the whole world, the US looks bad." Pellerin says that he realised he was the one person in the world that had the ability to salvage the Hubble project.
"I had a big budget for astrophysics programs. I had $2 billion or $3 billion a year to spend on various things. And I had the motivation and I had the team that knew how to do it because we built the thing in the first place," he says.
"So I quietly, and perhaps illegally, scrounged up $60 million and started a servicing mission. When I started we didn't know how to do it."
However, "it turned out that the nature of the error was a good thing," he says. The mirror was flawed but not in an irregular manner. "We built the perfect mirror to the wrong prescription."
His team worked out that removing one set of instruments — "one of them wasn't that important; it was a photometer" — they could insert another mirror in Hubble that was deformed proportionally to the telescope's flawed mirror, which would allow the half a wavefront error to be corrected by the time it hit the telescope's instruments.
"Once we had everything in hand and knew how to do it and showed people we knew how to do it, everybody forgot about the fact that I had started this thing against the wishes of the most powerful person in the world for NASA's budget. They ignored all that and nobody thought much about the leadership failure.
"So I actually got promoted twice to the top of NASA. I didn't like it up there. I like being close to hardware, scientists, technical problems. It is all politics up there; it's dealing with the White House, the Congress."
After 10 years of director of astrophysics, he decided to call it quits. It was time to do something else. "Probably a good idea to let somebody else do it anyway after 10 years," he says.
He decided to he wanted to get to grips with the concept of leadership; "Whatever it is, it trumped the best technical minds in the world." He got a professorship at the University of Colorado business school and began teaching a course called '21st Century Leadership'. As part of studying leadership, he took another look at the lessons that could be learned from the Challenger disaster and discovered a book by Diane Vaughan called The Challenger Launch Decision.
"What she said is, the real question is not the technical question. The real question is, why did they continue with the launch when all the data said they shouldn't? She said that there are social forces at play that are forever invisible and unmanageable. And it's most unfortunate she said… She named the phenomenon normalisation of deviance.
"It's for things that are deviant if you step back from them become okay locally. What happened with the Challenger launch is that under the pressure from Washington to launch, launch, launch, the technical people at Marshall [Space Flight Center] drifted unnoticeably into a place where it required a much more powerful technical argument to delay a launch than to continue. I looked at that and I said, 'By god it's the same thing that happened with Hubble.'"
As part of his study of leadership he began to look for other cases of the phenomenon. One powerful example he came across was the rash of crashes suffered by Korean Air Lines in the 1990s.
"Korean Air Lines in the 1990s was crashing at seventeen times the international average," Pellerin says. "It got so bad the president of Korea would not fly on Korean Air Lines. What's interesting about it is it went on for four years. Why did that happen?"