These numbers are not necessarily accurate. I'm still working on the collection method.
This is a chart of the number of console logins — that is, login sessions where the user was physically present at the computer — per lab per day. Since you can't log in to the windows lab remotely, this is the only kind of login session available there. This doesn't count any remote connections to the linux lab, of which there are many. It also doesn't count linux login sessions which lasted less than one minute.
This is the only reliable metric currently available for both labs. The only usage metrics collected from the windows machine at present are timestamps of when someone logged into a machine.