Wednesday evening I took my last D0 shift. I’m sure shifts have been covered here before, but we basically need people here 24 hours a day to make sure the detector is recording data, and fix any problems that come up. At D0, we have 4 people in the control room: 2 detector experts, 1 person making sure the collisions are being recorded properly, and 1 “captain” (the shifts I take) to make sure the good ship D0 stays on course. Or at least doesn’t stray too far off course.
My shift was 4pm to midnight, and frankly it was a little dull. But this is good – it means there were no problems, and we took lots of data. So at least I didn’t go out with a bang… Once a collision has been recorded, it is saved on tape at the Feynman Computing Centre here at Fermilab. Then, it gets moved to a computing farm to be “reconstructed”: turning all the 1s and 0s from the readout into more meaningful quantities, like the signal of a high energy particle. After reconstruction, it is ready to be analysed: I think the record for shortest time between a collision being recorded to appearing in an physics analysis that went public is just under 1 month. Occasionally we come across something unexpected when analysing a data sample, and have to trace it back to what was happening with the detector a few months or even years ago.
So we have to stay alert on shifts – not always easy at 4am on an overnight! If things really go wrong, we get the experts in: this is a picture from a much more exciting shift I had a few months ago.
