Monitor the fleet

This page describes how you keep an eye on your devices: where they are, whether they are online, what they report, which firmware they run, and how you are told when something goes wrong.

Inventory

The Hub lists every device of a registry in a searchable, filterable inventory. Each device has a detail page with its configuration, states, commands and actions, alerts, credentials and a remote terminal.

Gateways, sub-devices and beacons are first-class devices, each with its own serial number and page:

• a gateway connects to Fundamentum and reports on behalf of other devices;
• a sub-device is wired or paired to a gateway, and has no connection of its own;
• a beacon is seen over the air by a gateway, which reports what the beacon says along with its signal strength.

A dedicated view lists only the gateways of a registry.

Live connectivity and reboot detection

Devices send periodic heartbeats carrying their uptime, the time since they last started. Any other traffic from a device also counts as a sign of life.

• A device is connected as soon as it is heard from, and disconnected once it stays silent longer than the timeout set on its device type.
• When a device's uptime goes down between two heartbeats, Fundamentum knows it has rebooted.

The heartbeat history of each device is available in the Hub and through the Public API.

States

Devices report states: key/value observations such as a temperature, a door position or a firmware version. Fundamentum keeps their history and their latest values, and the Hub charts them per device. See Collect, route and transform data.

Firmware tracking

Each device type lists its firmware versions. Fundamentum records which version each device runs, and knows when a newer version is available for it, so you can target the right devices with a firmware update. See Operate remotely.

Alerts

An alert rule watches device states and raises an alert for each device that meets its conditions.

• Conditions. A rule has one to ten conditions, each comparing a state key with a threshold (=, !=, >, <). The alert is raised when all conditions are met, and ends on its own as soon as one of them is no longer met. You can also resolve an alert by hand.
• Hysteresis. For > and < conditions, you can set a second, release threshold: once triggered, the condition stays triggered until the value crosses that release threshold. This keeps a value hovering around its limit from raising and clearing alerts over and over.
• Six severity levels: info, advisory, warning, minor, major and critical.
• Scope. A rule applies to every device of its project, or only to the devices of one registry.
• Messages. A rule can carry its own message, in English and in French.
• Notifications. You choose who is notified, by user or by role. Recipients are notified by email and in-app; repeated occurrences of the same rule are grouped so recipients are not flooded with emails.
• As a data stream. Alerts are also available as a platform event stream that you can route to your own systems.

Go further

• Send telemetry, states and heartbeats
• Gateways and sub-devices
• Alerts
• Firmware updates (OTA)
• Topics: consume alert and device event streams.
• Public API
• MQTT topics and message formats