[Steve Collins] is a regular around Hackaday. He’s brought homebrew LIDARs to our regular meetups, he’s given a talk on a lifetime’s worth of hacking, and he is the owner of the most immaculate Hackaday t-shirt we’ve ever seen.
For the 2016 Hackaday SuperConference, [Steve] took a break from his day job of driving spacecraft around the Solar System. As you can imagine, NASA plans on things going wrong. How do you plan for that? [Steve] answers all your questions by telling you what happens when things go wrong in space.
Space is the worst possible place for hardware. Not only do you have temperature swings of hundreds of degrees, solar radiation, and limited bandwidth, but you also can’t fix a space probe once it’s in orbit. Anyway you look at it, everything needs to go perfectly or you need to be exceptionally clever. Muphry’s Law will inevitably crop up to defeat the former, leaving the latter par for the course. This is what it’s like to work at the Jet Propulsion Lab.
Most satellites that go up are, surprisingly, very standardized. GPS satellites are built around a two or three common ‘busses’, or models. When a company wants to launch a few dozen communications satellites, the first one-off the pad won’t be much different from the last. Whenever SpaceX gets around to launching four thousand of their low orbit Internet satellites, all of those birds are going to be the same.
Deep space satellites are completely different. Each one is a custom build, and the best examples of twin deep space probes – Spirit and Opportunity on Mars, and Voyager 1 and 2 – are the exception rather than the rule. A unique piece of hardware flying around the Solar System presents a few challenges for the hardware designers. Power is always an issue, you need to plan for redundancy, and every piece of hardware needs some sort of fault protection system. Everything is a challenge in designing a deep space probe, and you need to plan for every contingency.
This is the theory of designing hardware that has to work perfectly in the worst environment imaginable, but how about some practical examples of what to do when things go wrong in space?
Throughout [Steve]’s storied career, he’s been a part of a lot of NASA missions. In the 90s, one of his jobs was planning the Deep Space 1 mission. This was a mission to a comet done on the cheap — only about $150 Million – used to demonstrate up and coming technologies like ion propulsion. While in the planning stages, [Steve] and his colleagues discussed what could go wrong. Since this was a very inexpensive mission, only one star tracker was flown on this tiny satellite.
This star tracker is important, as it’s the only thing on the spacecraft that tells the computer where it’s pointing. In the planning stages, [Steve] discussed what would happen if that star tracker died. The hypothetical solution to this problem used the science camera to point at a single star and determine the probe’s orientation. This solution sat around in the back of [Steve]’s mind for a few years until — you guessed it — the star tracker died. It wasn’t pretty, but the hack of using a science camera to determine the spacecraft’s orientation worked.
That’s a sample of what happens when things go wrong in space. What happens when things go right? Check out the video below. That’s a car, landing on Mars, with the help of a rocket-powered crane. It’s Curiosity dropping into Gale crater, and [Steve] was in the control room for this astonishing feat of engineering. He’ll be doing it again in late 2020, and with this guy at the helm we shouldn’t have much to worry about.