Events Explorer
Pivot analysis of device events for a chosen corrector type — top codes by frequency, distribution across five classes, and per-node drill-down.
Events Explorer is a tool for free-form exploration of fleet device events for a chosen corrector type. It collects events from several nodes of the selected type and brings them into a single picture: which events happen more often, which categories they belong to, and on which nodes they occur. The audience is the control service and analysts: not a ready-made verdict, but a research tool to understand how a fleet of a specific device type behaves and which events are worth tracking.
What it shows
A summary of all events of the chosen device type over the period, broken down across several cross-sections.
| Element | What it means |
|---|---|
| Total events | how many events were collected on the selected nodes over the period |
| Unique codes | how many different kinds of event occurred |
| Most frequent codes | a list of events sorted by frequency, with the share of the total |
| Event categories | distribution across 5 groups (see below) |
| Breakdown by node | on which nodes the events occurred — you can drill down into details |
Each event is assigned to one of 5 categories:
| Category | What it means |
|---|---|
| Physical access | housing breach, device tampering |
| Substitution | substitution of data or metering mode |
| Metrology | measurement deviations |
| Communication | data-transmission failures |
| Other | the remaining service events |
How it is calculated
First, the list of fleet nodes is found for the chosen corrector type. Then, from a portion of these nodes (10 by default, to avoid overloading the platform) events are collected for the specified period from the archives of abnormal situations, changes and alarms. The collected events are grouped by code, category and archive kind, after which summary tables are built: the most frequent codes, the distribution by category and the breakdown by node.
Run parameters
| Parameter | Default | Meaning |
|---|---|---|
| Corrector type ID | required | the device kind to break down (for example, 2 — Flowgaz/ULTRAMAG, 7 — Type-J, 13 — EK270, 14 — Irvis); choose a type that keeps event archives |
| Period from | last 30 days | start of the event window |
| Period to | yesterday | end of the event window |
| Number of nodes to probe | 10 | sample size of nodes; more is more accurate, but slower |
How to read it
Start with the most frequent codes — they show what the fleet of the chosen type mostly raises noise about. Then look at the categories: a spike in physical-access or substitution events is a reason to look more closely at specific nodes via the breakdown by node. Metrology and communication more often indicate technical rather than deliberate causes.
Keep the limitations in mind. By default only 10 nodes are polled — for reliable statistics on a type with hundreds or thousands of devices, raise the number of polled nodes. If the chosen type keeps no event archives, the result will be empty with a warning — choose a type that has events.
Related reports
The Corrector Failure Atlas provides a reference of event codes, while the per-node Metering Bypass report places the events of a specific node on a time scale.
Related topics
Was this page helpful?
Thanks for your feedback!