You build the logic. RCH gives you the dashboard.
Self-hosted control dashboard for robotics and automation. 42 widgets, 4 protocols, zero frontend code. Docker to dashboard in 15 minutes.
Screenshots coming soon
Try the real thing ↑
42 Widgets. 6 Categories.
From joysticks and video streams to charts and data tables — ready-to-use components for any control scenario.
Control
Interactive widgets for sending commands to hardware — buttons, joysticks, sliders, number inputs.
Screenshots coming after widget polish
Modern Dashboard Interface
High-performance telemetry visualization with low latency.
Real-time Robot Control
Direct command execution with visual feedback loop.
Modular Widget System
Drag and drop interface to build your custom control plane.
The Problem: Why are we still rebuilding dashboards?
Backend and embedded engineers spend weeks building fragile custom dashboards — work they neither enjoy nor excel at. The alternative is no UI at all (terminal/SSH), which limits who can operate the system.
RCH eliminates this bottleneck entirely. Zero frontend code required. Developers focus on logic and automation; RCH handles the control surface.
Why Not Cloud?
Privacy
Your data stays on your local hardware.
Latency
Real-time control needs local speeds.
Offline
Works even when the internet doesn't.
What RCH Gives You
Connect your existing backends (MQTT, REST, WebSocket, ROS2), drag widgets onto pages, bind them to data — working control panel in under 15 minutes. No JavaScript. No CSS. No build tools.
You bring:
- Automation flows & Device logic
- ROS2, MQTT, REST or WebSocket endpoints
- Raw telemetry & sensor streams
RCH provides:
- 42 widget types
- 4 protocol connectors
- 13 data transforms
- ACK-confirmed commands
- Video & audio streaming
- Joystick & D-Pad
- RBAC & audit logs
- 13 languages + PWA
Who Is It For?
Built for the hands-on engineer and hobbyist who needs professional-grade control.
Robotics Engineers
Drag a Joystick widget, bind it to /cmd_vel — your robot moves. No frontend code. ROS2 and MQTT out of the box.
Factory & Lab Automation
ACK-confirmed commands with 2s timeout. Operators press "Emergency Stop" — you know it executed. Audit log tracks who did what.
R&D Teams
Prototype a control panel in 15 minutes. 42 widgets, 13 data transforms, 4 protocols. Iterate on logic, not UI.
Makers & Educators
Students control TurtleBots through a browser instead of terminal. 30+ concurrent users on a Raspberry Pi. MIT licensed.
AI-Powered Setup (MCP)
Built into the image — no separate install. Connect Kiro, Claude, Cursor, Windsurf, or Continue.dev and build dashboards through natural language. 37 tools for widget creation, data binding, and source management.
Transformation Engine
13 chainable transforms: low-pass filtering, scaling, deadzones, clamping, mapping, and more. Adapt raw sensor data without writing code.
4 Protocol Connectors
MQTT, REST, WebSocket, and ROS2 in one tool. Widgets stay logic-free — connected via declarative bindings with JSONPath data mapping.
ACK-Confirmed Execution
Three command modes: fire (instant), ack (confirmed with 2s timeout), submit (buffered). Know the real state of hardware, not just that a command was sent.
See It Live
A fully functional alpha — 42 widget types, 4 protocol connectors, real-time WebSocket.
One Command. Full Control Panel.
RCH ships as a single Docker image — FastAPI backend, Vue 3 frontend, PostgreSQL, and Redis. Runs on a Raspberry Pi, your laptop, or a server. From docker compose up to a working dashboard in under 15 minutes.
Starts in 30 Seconds
Single docker compose up — no build steps, no dependencies, no configuration files.
Your Data Never Leaves
Self-hosted, no telemetry, no cloud dependency. Runs entirely on your local network.
Runs Everywhere
Jetson Orin, Raspberry Pi, laptop, or cloud server. Native ARM64 and AMD64 images — no emulation.
MIT Licensed
Free and open-source core. Use it commercially, modify it, contribute back.
Solo Developer Transparency
RCH is currently a solo project. It's built out of a personal need for a better way to control local devices without relying on external clouds.
Stop building dashboards.
Start controlling hardware.
42 widgets. 4 protocols. Zero frontend code. One Docker command.
Try the live demo, deploy on your Raspberry Pi, or tell us what you need.