Creating Rules
Create a new automation rule from scratch in the Kilo IoT Rules Engine — name it, design the workflow, save first.
This page walks through creating a new automation rule from scratch — from naming it to saving your first version.
Prerequisites
You have edit permissions for the Rules Engine in your organization
Your organization has not reached its subscription rule limit
Step by step
1. Open the rule creation page
From the Rules Engine page, click Add Rule. This opens a new rule editor at /rules/create, with an empty name field and a starter diagram — a Start event already connected to an End event.
2. Name your rule
The rule name field is at the top of the editor. Click it and type a descriptive name — up to 60 characters. Choose a name that tells your team what the rule monitors and why.
Good naming conventions for operations:
"Cold storage — temperature breach"
"Warehouse B — humidity alert"
"CO2 monitor — threshold escalation"
3. Add a description (optional)
Click the three-dot menu (⋮) next to the rule name and select Edit description. A dialog opens with a text area — up to 1,000 characters. Use this to explain the rule's purpose, which facility or zone it covers, or any operational context your team should know.
4. Build the workflow
The editor opens with a Start event and an End event already on the canvas. A complete rule needs at least:
A Start Event — the entry point that uses either one sensor's readings or a saved trigger condition
One or more processing nodes (Script Tasks, Gateways, Enrichment, Set Alarm)
An End Event — at least one termination point
Drag nodes from the palette on the left side of the canvas. Connect them with flows (arrows) to define the execution order.
For many rules, this visual structure is the main work: Start Event, one or two decision or transformation steps, then Set Alarm or End Event. When the rule needs exact logic, specific fields accept CEL expressions — for example gateway conditions, Script Task logic, alarm motivation messages, or advanced inputs/outputs. Through CEL, even complex scenarios are possible: multi-step escalation, cross-sensor comparisons, computed derived values, and dynamic alert messages that include live readings.
For a detailed walkthrough of the canvas and available tools, see Visual Editor. For the full list of node types and their configuration, see Node Reference.
5. Configure the Start Event
Select the Start Event node and click the pencil icon that appears beneath it — that opens its properties panel on the right. First choose what starts the rule:
Start source — Sensor reading runs the rule whenever one selected sensor reports. Trigger condition runs it when a saved trigger becomes active. A trigger can act immediately or after a duration, and it can evaluate one device or several devices independently.
A Start Event uses one source or the other, never both.
With Start source set to Sensor reading, configure:
Device — Select the device that triggers this rule (searchable dropdown)
Sensor — Select which sensor on that device to monitor (dropdown, enabled after choosing a device)
With Start source set to Trigger condition, the Device and Sensor fields are replaced by a single Trigger condition selector. Pick the trigger you built on the Triggers tab. This is the path for a condition that must be evaluated before the rule starts, whether that condition is immediate or delayed and whether it watches one device or many. The trigger signal does not contain one sensor-event value, so vars.value is unavailable; vars.device_name identifies the watched device that met the condition.
The selector does not create a trigger. If the trigger does not exist yet, leave the editor, open Rules Engine → Triggers, create it, and then return to this rule. After selecting it, click Save at the bottom of the Start Event panel to apply the source to the diagram.
Optionally, you can:
Enable Schedule — Toggle to restrict the rule to a specific window. Click Change schedule to pick the days and the From/To times, and set the Time Zone they are measured in. Schedule is not another start source: it limits when the source you already chose may run the rule. With a duration trigger, the trigger decides when its condition becomes active and the schedule decides whether the rule may run at that time.
Add inputs/outputs — Advanced CEL expressions for data transformation at the Start Event
See Node Reference for full Start Event configuration details.
6. Save your rule
After saving the Start Event panel, click the separate Save button in the top-right corner of the editor. The rule is saved as a new version (version 1). The autosave indicator shows "Saved" when the save completes.
After saving, the editor switches from creation mode to edit mode. You now have access to the History tab, the Build button, and the full actions menu.
7. Build and deploy (when ready)
Saving a rule does not deploy it. To run the rule against live sensor data, you need to build and deploy it. See Builds, Artifacts, and Deployment for the deployment workflow.
What happens after creation
The rule appears in the Rules tab of the Rules Engine page
Version 1 is recorded in the version history
You hold an edit lock on the rule — others will see the lock icon
Autosave is active — changes are saved periodically while you work
The rule is not running until you explicitly build and deploy it
Next steps
Visual Editor — Learn the canvas tools, palette, and properties panel
Node Reference — See all available node types and how to configure them
Builds, Artifacts, and Deployment — Build and deploy your rule to production
Last updated