Digital Aquatics Reef Keeper controller application data science?

[dead_goby]

I've been developing an add-on for Splunk to manage and report on data being generated from the DA NET module.

I have a couple of goals here:

1) Track stability of metrics, the assumption is that reef tanks thrive on stability, having these metrics will help me see how things are moving over time.
2) Use machine learning to detect faults before they become serious, using the predict command we detect a fault like a stuck heater before it becomes terminal.

I'm not a data scientist so I'm open to feedback or criticism. If you have any ideas on how to use the machine data for being analytics, reporting and fault management please let me know.

 

[StlSalt]

This is interesting. Following. Never heard of Splunk until now.

 

[Azure Scolymia]

One of the things that you will have to consider are variations that are expected vs variations that are not expected. Take pH for example. You get pH changes that are associated with the lights turning off and on. That is probably not an issue, but will complicate your ability to detect smaller, but real variations.