The UM Computer Aided Engineering Network (CAEN) runs engineering computer labs across UM. CAEN has screens outside each lab showing how many computers are in use. They’re just running a webpage, so I dug into the source code. It gets its data from a JSON file.
I wrote a PowerShell script to get the JSON file and count up the total computers and computers in use from every lab. The result was outputted to a CSV file. I ran the script as a background job once a minute for 15 days with a gap of 4 days when the computer was turned off.
Each peak is one day. The days starting in the middle are Tuesday-Monday. In the last couple days, the script recorded data irregularly: usually 5-10 minutes between data points.
The usage declines throughout the week. Perhaps it’s because people get lazier? Saturday was much lower than the other days, but then people get back to work on Sunday: it’s almost as busy as a weekday.
This is September 29th, a Thursday.
What’s interesting in comparing the days side by side is that Friday drops off much more quickly than the other weekdays between 4:00 PM and 6:30 PM. There’s Friday for you. Sunday gets going about an hour later than the rest of the days. For instance, at 10:31 AM on Sunday, 151 computers are in use. The next latest to get to that point is Thursday, at 9:39 AM. This was just one week of data so this may not be how it always is, but it’s probably close.
At around 8 AM every day, there’s a sharp drop in usage. It may be that people who left themselves logged in for a long time are automatically logged off.
You can run the script without any modifications. By default, it will save the CSV to C:\Temp.