Wednesday February 25^th 2009. Launch -50 days.

This is turning out to be an extremely tense and tiring week, but not in the way that I expected initially. The last two days have involved seemingly endless, highly technical brainstorming sessions where we have discussed how to handle the up-coming second simulations campaign. This is our last chance to test all the computers and software that we will be using at launch and relies on many different elements being ready at once. The upshot of this is that the Test Plan that I have to modify for the test campaign needs to be very extensively re-written. Given that it is a highly technical document, nearly 150 pages long and was intended to be ready on Friday (yes, that’s the day after tomorrow), there is not a hope in hell of me being able to meet this deadline, even if I were able to dedicate most of my time to the task (which I cannot). Today I have had two long, tough meetings, particularly the one in the afternoon with the instrument teams where there was what is normally called “some straight talking”. In the end, that meeting finished well. Why? Well, it all came down to people taking a positive and constructive attitude and doing their best to find solutions to difficult problems. Every launch campaign turns out to be a race against time, with a final sprint to the line and ours is going to be no different. As marathon runners say: it’s not the 42km that hurt, it’s those wretched extra 185 metres at the end!

Tomorrow morning will be taken up by a long meeting in the morning that will probably last almost until lunch, so no work will be done on the Test Plan then either, but the meeting is absolutely vital to agree on the ways that the instrument teams will deliver their observations to test the instruments after launch: in a telescope that is 1.5 million kilometres away and operating automatically most of the time, with no direct control from Earth and following instructions placed in its on-board memory, everything has to be supplied spoon-fed in tiny baby steps – computers are very literal creatures and have no imagination, hence they do exactly what they are programmed to do, right or wrong (hence the standard explanation when someone receives an electricity bill for an entire city that it has been a “computer error” – computers do not commit errors, but the people who program them do commit some beauties!) What this means is that any mistake that we make in our instructions to the on-board computer may come back to haunt us later. The solution is to try to eliminate human error completely: very carefully honed step-by-step procedures, precisely defined agreements on how we are going to name files, prepare them and process them, etc. It’s a pedant’s dream, but a human nightmare because there is a danger of going too far in trying to make everything foolproof and make everything so complicated that people commit errors trying to avoid committing errors.

For the second consecutive evening I have fallen asleep in my chair with a Harry Potter film in the DVD. And to think that this week is the easy week: next week is going to be really tough.