Skip to main content
Automations let you connect device events, conditions, schedules, and actions into a workflow. Use them when a device should react automatically without an operator watching the dashboard all day. [image:automation editor with device trigger, condition, action, schedule, and execution trace]

​
What automations are for

Use automations for workflows such as:
  • Turn on a relay when a sensor value reaches a threshold.
  • Send a command to a device on a schedule.
  • Check multiple conditions before running an action.
  • Run different actions depending on device state.
  • Use AI to classify, decide, or process information before continuing.
  • Stop or continue a device workflow based on an execution result.

​
Main views

The Automations page has two main views: An automation can be active or inactive. Inactive automations stay saved but should not run until they are enabled again.

​
Automation editor

The editor is a canvas where you connect nodes. Each node has a role in the workflow. [image:automation node picker grouped by device, logic, schedule, flow, and AI]

​
Schedules

Schedules can run an automation without waiting for a device event. The current editor supports common repeat choices such as daily, weekly, monthly, yearly, once, and custom date selections. For scheduled automations, choose the time and timezone carefully. If a schedule depends on a specific date, day of week, or day of month, review it in the calendar view before relying on it.

​
Testing and traces

Use Test Automation before relying on a workflow in production. The editor can listen for live execution events and show which steps ran, which path was taken, and whether the result succeeded or failed. The Execution trace tab shows past runs for an automation. It helps answer:
  • Did the automation run?
  • Which node ran first?
  • Which condition path was selected?
  • Did any step return an error?
  • What was the final result?
[image:execution trace timeline showing successful and failed automation steps]

​
AI and knowledge bases

Some automation nodes can use AI. AI nodes can classify information, make decisions, process images, or run an agent-style task. When knowledge bases are selected, the AI node can use uploaded knowledge as context. Use AI nodes when rule-based logic is not enough. For deterministic device safety rules, prefer clear conditions and device actions.

​
Safe usage checklist

Before enabling an automation:
  • Confirm every device and entity is selected correctly.
  • Make sure the first trigger is intentional.
  • Check schedule timezone if the automation runs by time.
  • Test the automation and review the trace.
  • Keep the automation inactive until the result is expected.
  • Use clear names and descriptions so other workspace members understand the workflow.
Automations are workspace-scoped. A user should only see and manage automations available to their authenticated workspace context.