The Galileo probe and its cold welding problems
I wanted to get a little further into detail on the problem with the Galileo probe, which spent seven years orbiting Jupiter and would be the last Jupiter orbiter until Juno in 2016.
Even before launch, Galileo had drama. It's powered by a plutonium generator, and there was fear of it being hijacked. So the truck driver wasn't told the route ahead of time, and they drove all night and day, only stopping for food and fuel.
And then when it was in Florida, ready for launch, the Challenger disaster happened. It had to be shipped back to JPL in California. In all, Galileo was in storage for nearly a decade before it would launch in 1989.
This meant the critical high-gain antenna's lubricant was worn out from age and vibrations in transit even before getting to space. On trying to deploy it along the way to Jupiter, the signal indicating success never came.
While the antenna had been tested before launch, the cold welding problem wasn't replicated. While they tested deployment in a vacuum, it didn't include all of the factors that lead to failure, including the vibration during launch that caused the parts to rub together (fretting).
NASA tried for a while to get the antenna to deploy. The first idea was the classic 'turn it off and back on again' by folding the antenna back up. But while the motors could reverse, the mesh antenna was found to tangle when tested on Earth. Eventually they concluded it really was stuck and they'd have to use the low-gain antenna, which doesn't concentrate its signal toward Earth.
So how do you do science when transmitting data at less than one ten-thousandth of the desired speed? They went from 134 kilobits per second to 8-16 bits per second. Data compression! Gerry Snyder was an engineer who worked on Galileo in the 1970s (did I mention Galileo sat around forever before launch?) and helped get the data transmission working.
Compounding the slow transmission problem was that Galileo had a limited window to send back data. When it's orbiting the back of Jupiter, it can't contact Earth. Galileo had to be reprogrammed to limit what it sends back. Duplicated data was filtered out, and everything else combined into packets which were only sent when there was new information.
Also vital to the mission's success were improvements to the Deep Space Network. The low-gain antenna, not being directional, was a bare whisper in radio signal on Earth, and picking it up and filtering out the noise was a mess.
Ultimately, Galileo still managed to complete some 70% of its scientific goals. It performed the first-ever flyby of an asteroid, and it was able to watch the impact of Comet Shoemaker-Levy 9, fragments of which created a 1,900 mile high fireball. It was the only instrument to directly observe the impact, since the site was facing away from Earth at the time.
It also carried a separate atmospheric probe, released 50 million miles from Jupiter, which managed to send data for almost an hour while descending through Jupiter. The Galileo probe was so successful a follow-on project had to be set up for even more scientific investigation, until the probe finally fell into Jupiter in 2003. This an intentional action to prevent it contaminating Galileo's own discovery of Europa's possible oceans.