ARDI Trigger
Trigger is a small Python library that allows you to create triggered responses. These responses can be based on live ARDI data, or the current time.
It allows you to do practically anything in response to changes or patterns in your live data. While Agency is designed to produce alerts, Trigger is designed to produce responses to events, particularly when they cover ranges of time.
Trigger functions can be run across simple, pre-determined spans of time - such as daily summaries, end-of-shift totals or weekly KPIs.
But by using your data as a trigger, you're able to perform analytics on dynamic ranges of time. These include summaries per batch, crew, production run or product type. You can also perform analysis of individual machine actions, such as startups and shutdowns.
Benefits
Trigger has the following advantages…
- Manages the subscription to live ARDI data,
- Allows you to define custom, complex logic for each trigger
- Allows control of trigger edges and delays to avoid excessive activations,
- Launches each trigger event in its own thread or process, ensuring they don't interfere with one-another.
- Avoids overlapping jobs - a single trigger won't be run again until its job completes. Changes during computation are deferred until computation is complete.
- Manages simple key-value and event output
- Offers a backfilling mechanism to apply new analytics across previous time-frames
- Reduces calls to the condition function as much as possible.
Outputs
Trigger does provide access to the Modular Output System for key-value and event data.
But because Trigger functions are written in Python, you have flexibility to use any type of output that you like. If you want to write to a database, FTP a file, send an SMS notification, produce a PDF report, make a PA announcement, order pizzas or create a video file, Python provides the tools to do so.