An integral offering of any IoT-driven monitoring solution is capturing, processingand generation alarm conditions to users. For example in case of Fleet Management, an alert could be a vehicle going over a speed limit, or in case of Quick Service Restaurants (QSR) it could be an alarm condition generated because of rise in the freezer temperature.
These alarms or events are valuable tools for near realtime operational functions, if the freezer temperature is rising – quick corrective action would lead to ensuring food safety and preventing loss of inventory. However typically once the alarm is generated and captured in a report, it goes away as a distant memory. Yes, the system will probably have the temperature values recorded with some knowledge of an alert conditioning occurring.
The optimal approach, as we do in our platform, is to treat every single alert(quantitative or qualitative), alarm condition, and breach of threshold values along with recipients of the corresponding notification as an “event”. Any event (based on especially alarm conditions) is stored as a part of time series data on events.
What are the benefits?
- An event-only (i.e. without the data that was the source of these events) time series that can be analyzed in itself – as its own source of information. This, we believe, improves the source of ‘labeled’ data for AI-driven predictive maintenance.
- Playback – Events, if the app is designed correctly, can be played back and understand what relationship it has with the system.
- Auditing – for reasons ranging from compliance to simulate alert conditions – the log of events serve as an audit trail.
Applications designed using the iMsight IoT Application Framework gain from this intrinsic functionality.