Hi Steve,
One of the biggest challenges in my work is that sometimes scientific instruments on rockets and satellites do not work properly. We do a lot of testing on the ground to make sure our instruments are calibrated and that we understand their limitations. For example, we have a vacuum chamber here in one of our laboratories that we use to test electron detectors for NASA sounding rockets. We can also do computer simulations of how our electron detectors work. One of the things that I am working on now is simulating the response of the Electron Drift Instrument (EDI) for the upcoming Magnetospheric Multiscale Mission (MMS). Some instruments are difficult to test under the exact same conditions that we will find in space, such as the huge antennas we often use to measure plasma waves, so we do our calibrations with the electronics packages and carefully analyze the data after the satellite is launched to look for any discrepancies with what we expected to see based on our tests on the ground.
Unfortunately, even when we do all of these tests and simulations, things can still go wrong. With a few exceptions, we can't just fly up there and fix problems with a satellite or rocket after it is launched. On the sounding rockets I am currently working with, there were problems with several instruments due to changes in one key part that were not well-documented by the manufacturer. As a result, the instruments did not all work as planned. Sounding rockets are also put together very quickly and often used to test new instrument designs that have not been tried in space before. It is very frustrating and disappointing to put a lot of work into a space flight instrument and discover that something does not work properly. In some cases, the data are totally useless, if any data are returned at all. In other cases, we can recover useful information from the data, but it takes a lot of extra work to compensate for the problems. Space exploration is risky, so when things go wrong, you do your best to evaluate and document problems so that the instrument can be redesigned for the next mission.
Kris