Problem confirmed, problem (almost) solved

In the modern society where technology is pervasive and even the simplest devices are full of different parts and even run their own software, one of the most haunting situation you can face is an intermittent failure. It is not difficult to explain and understand that, as things get used, they start to show signs of wear and they eventually do not work as well as they used to or they stop working whatsoever. This is a predictable phenomenon and, save for small uncertainties as to how fast the wear will ensue, we understand it an expected.

The problem arises when problems "disappear": the universe is not supposed to work like that, things get worse with time, one would never expect a smashed egg to reconstruct itself, or a jigsaw puzzle coming assembled out of the box, but in modern life some problems seem to go away on its own. Of course, the deeper explanation is that the problem requires a relatively rare combination of seemingly unrelated factors to happen at the same time before it can manifest itself. If you accidentally remove one of the factors, the problem does not show anymore, but it is far from solved. Restore the the whole set of conditions that led to its appearance and, sure enough, you will see it again. I have seen this many times both in my professional and personal life and the task of fixing the problem can be daunting: your best hope for solution is to be able to induce the problem at will, so that you can see it happen time and time again, observing which factors contribute to the problem and which ones not, guiding you to the eventual solution of the problem.

Photo: PxHere

About two months ago we received an image from our instrument that showed a salt-and-pepper pattern in some portion of it. The image had been processed on board the instrument, then compressed and finally transmitted to ground, but we did not have a clear idea where the failure could come from. It was just one of several images, and the others were correct. We asked the ground station to send that image again and the pattern was still there, so it obviously happened before the data hit the ground. It could be due to the compression, the on-board processing, or even the acquisition, so we had to trace our steps back.

To be successful in space science the most essential quality is patience. The construction of space instruments is a very lengthy process, traveling through space takes a very long time, and even the reception of data from the spacecraft is carefully metered, so you better be prepared to wait for long periods of time before you embark on such a mission. Yesterday we finally received the uncompressed version of the image that showed this funky pattern and it was already there, so the compression is not the culprit either. Now the hunt continues, but always at snail speed, more like tracking Sasquatch in the forest of Canada than a car chase in the streets of Paris. Hopefully, we will eventually get the error and fix it. I will keep you posted, have a nice evening.