I had no idea this was out there -- and I feel like I closely monitor this group of brilliant people for the cool stuff they give our community! You can check out the iMonDB manuscript at JPR here.
After running into some of these authors I realized this needs to be backdated here!
It is the other side of the QC/QA issue that I don't think we consider that much -- what are the physical properties of the instrument itself telling us about it's ability to operate at peak performance right now?
As amazingly well as mass spectrometers are manufactured these days it may be safe to say that a few might not successfully run 24-7 until our sun reaches it's red giant stage ~5.5 billion years, when we we'll have more concerns than a few minutes of downtime.
iMonDB is free software available here that extracts data out of instrument raw files that most of us never look. Things behind the scenes you can find deep in the scan headers that one at a time, probably don't tell you much -- but plotted over time and correlated with rare instrument events can be powerful predictors of an upcoming repair -- before that component actually fails!
Sounds like a dream, right? This group has hard data to show that it does work. For anyone in the field, this could be a massive increase in convenience and up-time. For those of us trying to get our technologies into clinical settings...it doesn't take much interpretation of those requirements (good ref here) to think this is or will be a requirement for those settings.