I'm glad and sad that that's over

No computers; high technology of a different sort today.

Shuttle Atlantis has safely returned to earth for the last time; the Space Shuttle Program is over. Which both makes me sad, and makes me breathe a huge sigh of relief. Thank goodness that is finally over.

As you might have gathered, I am highly ambivalent about the Shuttle Program. It is unique in that it just might be the biggest engineering success and the biggest engineering failure of all time, at the same time. What a contradiction!

On the success side: the Shuttle is by any measure a stunning human achievement and everyone who worked on designing, building, maintaining and flying Enterprise, Columbia, Challenger, Discovery, Atlantis and Endeavour has every right to be proud. I mean, we're talking about  a vehicle that:

  • takes off as a rocket, flies as a spaceship and lands as an airplane.
  • has an main engine (those cones on the back of the orbiter) with the highest power-to-weight ratio of any engine, period. Imagine if every pound of your car's engine produced a hundred horsepower!
  • achieves temperatures inside the engine that would not just melt, but actually boil iron.
  • orbits the entire planet in about 90 minutes.

and so on. It is an astonishing vehicle and one of the great achievements of our civilization.

But all that said, as much as it is an amazing engineering success, the Shuttle Program failed to achieve almost all of the important goals originally envisioned. Those goals included:

  • Frequent launches. The original vision was for weekly launches, then that was decreased to twice a month, then once a month. In practice, the average launch rate over the last 30 years was over three months between launches: between a third and a tenth of the originally envisioned launch rate. The launch rate is an order of magnitude lower than envisioned.

  • A cheaply reusable launch vehicle. Reusable? Absolutely, but the whole point of "reusable" was to lower its cost. In practice, the costs of inspecting the tiles, overhauling the engines, and so on mean that launching the Shuttle is actually more expensive than just building a new throwaway rocket. When you consider just the per-launch costs, it cost about $18000 per kilogram to get an object into orbit with the Shuttle. When you consider the total cost of the 30+ year program divided by the number of kilograms it has put in orbit, it was a whopping $60000 per kilogram. The program goal was to achieve costs of $1400 per kilogram in 2011 dollars; the Shuttle cost is orders of magnitude higher than envisioned. (*)

  • A reliable vehicle. I clearly remember that day in eighth grade, heading back to my locker to get my lunch when a girl in my home room class named Christine said in passing "Hey Eric, the Space Shuttle just exploded." I thought she was making an unfunny joke, and was horrified to learn moments later that no, she was absolutely right.

    During the enquiry that followed the Challenger disaster, Richard Feynman estimated that the failure rate would be about 1%, and he was spot on: 135 missions, two total-loss-of-vehicle disasters. Yes, space travel is inherently dangerous, but a ~1% chance of total loss of vehicle and crew is absolutely unacceptable. The safety of the Shuttle is orders of magnitude less than envisioned. (**)


Ultimately the only truly failed experiment is the one you learned nothing from; we have learned a tremendous amount from the Shuttle Program in both its successes and its failures, and so it was a successful experiment. It is my fervent wish that this knowledge not be lost. I believe that it is within our reach as a species to explore space; it's a big universe waiting out there and we can and should go see what's out there, both in person and with awesome robots.

I am genuinely awed by the engineering achievement that was the Shuttle Program. What's even more amazing than that is: we can and will do better, at lower cost and with greater reliability.


I'm heading off to Europe for a week to attend a wedding, and then spending a couple of weeks with my family and friends in Ontario. Unlike some previous years I haven't prepared automated postings for my absence, so the blog is going to go dark for a month or so. I hope you all have a pleasant August and we'll pick up with more fabulous adventures in late August or early September.


(*) Another unfortunate aspect of the Shuttle Program's massive cost is that money has been spent on the Shuttle Program getting people into low earth orbit that could have been spent on getting more awesome robots to other planets. I love having awesome robots on other planets. The scientific knowledge gained per dollar spent is way higher for robots on other planets than for humans in low earth orbit.

(**) The fundamental problem here is not so much technical as it is cultural. NASA management seems to have the curious and obviously false belief that the fact you didn't fall off a cliff yesterday is proof that it is safe to walk even closer to the edge tomorrow, even if you're not precisely clear on where the edge is. That is actually an algorithm for ensuring that you always walk off a cliff. The truly awful thing is that they walked off the same cliff in the same way twice.