Background Caching System
Most of the Visual Analytics work at high data resolutions - to make sure nothing is missed, most analytics use data points that are no more than 1 second apart.
While data at this scale works well when you're looking at 10, 20 or 30 minutes of information, you can notice that performance quickly drops when you try to look at longer time-frames.
After all, a full day contains 86,000 points of data - and the more channels of data you have, the more this is multiplied. This quickly makes daily reporting slow, and weekly or monthly reporting - which requires literally millions of points of data - practically impossible.
Introducing Background Caching
If you provide a database for ARDI to use, it can create a background cache of the analytic results it has built.
This means that you've got an easily accessible, extremely fast record of all of the events your analytic has detected. This allows queries to be run extremely quickly on any time-frame that the analytic has examined in the past.
Example
Let's imagine you created a pivot analytic calculating how efficiently each of your shifts/crews performed. You deploy it as an hourly dashboard, and as a daily report.
Every time your hourly dashboard updates, it adds any new information to the cache. That way when the daily report rolls around, the data is already there - the report takes only a tiny fraction of the time to generate.
When it comes time to do a weekly report or a monthly summary, ARDI can do the same - it gathers the results in moments instead of minutes.
Supported Modules
Caching isn't suitable for every analytic - it's available on the modules below…