DEVELOPERS
Run a SCADA with an AI agent on your laptop in ten minutes.
The Faceplate stack ships as Docker Compose: Studio, Runtime, the historian and an OpenClaw agent pre-configured for power grids. One command, then ask it to find every transformer in substation SUB-2.

Quick start
# 1. Get the agent configuration and skills $ mkdir -p faceplate_stack/openclaw && cd faceplate_stack $ git clone https://github.com/faceplatekirill/fp_openclaw.git openclaw # 2. Volumes for the SCADA database and logs $ mkdir -p faceplate/DB faceplate/logs # 3. Start Faceplate and the agent $ docker compose up -d # Studio http://localhost:9000/fp/studio # Runtime http://localhost:9000/fp/runtime # Agent http://localhost:18789
Then import the light demo project, a power-grid model with a simulator, and start asking questions.
What the agent does out of the box
Object discovery
Find any SCADA object in plain language. No tag paths to memorise.
“Find all transformers in substation SUB-2”Time-series analytics
Min, max, averages and integrals over any period, without building queries.
Data health audit
Frozen sensors, telemetry gaps and signal health in one diagnostic view.
Alarm intelligence
Bad actors and repeating issues across a shift.
“Top 3 recurring alarms in SUB-3 last night?”Four ways to start
Pick the depth you need.
| Option | What it includes | Who it is for |
|---|---|---|
| Foundation Docker Compose | The stack with an empty project | Developers deploying or building their own project |
| Light demo | Power-grid project with a simple simulator, about 5% live telemetry | Exploring the UI and the grid knowledge base |
| Data deep-dive | Project with a historical archive database | Testing history and trend skills at scale |
| Ready-to-go VM | SCADA, agent, simulator and archives pre-installed | Evaluation without any setup |
Local, on-premise models for production and closed networks; cloud APIs for open test projects. Integration support: openclaw@faceplate.io. Or open a support request →