Alerts & thresholds

An alert fires when a device reading crosses a limit you’ve defined. Those limits are thresholds, and you manage reusable sets of them as threshold packs. This page is about setting those limits and handling alerts as a user.

Configure thresholds with threshold packs

A threshold pack is a reusable bundle of threshold rules you can apply across devices, so you define a limit once rather than per device. Threshold packs live under Settings → Threshold Packs (admin-only).

To create one:

1. Go to Settings → Threshold Packs and select New Threshold Pack.
2. Give the pack a name and add rules. Each rule describes a metric, the limit it must stay within, and how serious it is if the limit is crossed.
3. Save the pack, then apply it to the devices or device groups that should use it.

A device’s currently active limits are always visible on its detail screen under the Thresholds tab (see Devices).

How alerts get routed

When a threshold is crossed, IoT Pulse decides who hears about it using the responder Group that covers the device:

1. The device belongs to one or more device groups and responder groups.
2. The responder group’s Shifts decide who is on call at that moment.
3. The alert is delivered over the group’s notification Channels — Email, In-app, or both.

So if you want to change who gets alerted, you change the responder Group and its shift — not the threshold (see Device groups & shifts).

Why a repeat alert can go quiet

IoT Pulse deliberately avoids spamming you with the same alert over and over. It applies a cooldown after an alert fires, so an issue that’s still ongoing won’t re-notify you every few seconds. There are two independent cooldown layers working together as a safety net — so even if one is misconfigured, you’re still protected from a flood.

Note

If you expected an alert and didn’t get one, a cooldown from a recent earlier alert for the same device is a common reason. The condition is still being tracked — it just isn’t re-notifying yet.

Acknowledge and handle an alert

Alerts you can act on become tickets — that’s the work surface for resolving a problem. When an alert fires:

1. Open the matching Ticket (from the Tickets screen, or from the in-app notification).
2. Assign it, set its status as you work it, and add notes.
3. Resolve and close it once the underlying issue is fixed.

See Tickets for the full lifecycle.

Tip

The AI agent also raises tickets on its own when it spots a developing problem — not every ticket comes from a threshold alert. See AI analysis.