Hi everyone!
I’d like to share my ESPome Vortex project - a smart controller that replaces the original PCB in Vortex BWO 155 hot water circulation pumps, transforming them into intelligent, self-learning devices fully integrated with Home Assistant.
What It Does
The Vortex controller learns your hot water usage patterns over time and automatically runs the pump just before you need hot water - no more waiting for hot water at the tap, and no more wasting energy running the pump 24/7!
Fully integrated with Home Assistant through the native ESPHome API - all sensors, controls, and learning data are automatically exposed as entities you can use in automations, dashboards, and scripts.
Key Features
Machine Learning - Builds a weekly usage pattern (7 days × 48 half-hourly slots) that adapts to your household’s schedule
Smart Water Draw Detection - Monitors outlet/return temperature differentials to detect actual water usage and trigger the pump
Disinfection Cycle Detection - Automatically recognizes when your boiler runs its legionella prevention cycle and keeps the pump running appropriately
Vacation Mode - Automatically enters low-power mode after 24h without water usage
Anti-Stagnation Protection - Runs the pump briefly every 48 hours to prevent mechanical seizure
Energy Monitoring - Tracks energy consumption per cycle (kWh) based on flow rate and temperature differential
Web Dashboard - Beautiful heatmap visualization showing learned usage patterns with multiple themes
Hardware
- ESP32-C6 microcontroller
- Dallas DS18B20 temperature sensors for outlet/return pipes
- Logic control for the pump motor (simple digital on/off signal)
- USB-C Power Delivery 3.0 power supply for the controller & pump
- Dual LEDs for status indication (learning enabled, pump running/manual mode, WiFi connection status)
- Hardware button for manual control and learning matrix reset
The PCB fits perfectly on the original Vortex BWO 155 pump carrier, making installation a drop-in replacement.
Home Assistant Integration
Once connected to your network, the device automatically appears in Home Assistant with all entities ready to use:
- Temperature sensors (outlet/return pipes)
- Pump control switch with state feedback
- Energy monitoring (last cycle kWh and duration)
- Learning status indicators
- Vacation mode detection
- Pump trigger type (manual, scheduled, water draw, etc.)
- Configuration controls (enable/disable learning, pump operation)
Create powerful automations like:
- Notify when pump runs unexpectedly (potential leak detection)
- Disable automatic operation when away from home
- Track energy usage trends over time
- Integrate with other home automation scenarios
Screenshots
Hardware
[Include 2-3 of your hardware photos here:
- PCB front/back close-up to show the hardware quality
- PCB mounted in the pump showing the professional integration
- The pump installed on pipes to show real-world application]
Web Dashboard
- The heatmap visualization with learned usage patterns
Home Assistant Integration
- Real-time temperature readings
- Energy consumption statistics
- Control buttons and status indicators]
Current Status & Looking for Collaborators!
The project is fully functional and has been running reliably in my home. All core features are implemented including:
- Learning matrix persistence in flash memory
- Web UI with real-time heatmap visualization
- Configurable thresholds via YAML
- Home Assistant integration via ESPHome API
I’m looking for people who want to participate in this project! If you have a Vortex BWO 155 pump and want to upgrade it to a smart, learning system, I’m willing to sell PCBs to interested community members. This is a great opportunity to:
- Get involved in an active ESPHome hardware project
- Help test and improve the system in different environments
- Contribute to documentation and feature development
- Join a community of smart home enthusiasts working on practical solutions
If you’re interested in purchasing a PCB or collaborating on the project, please reach out!
Repository & Documentation
The complete project (ESPHome component code, web UI) will be published on GitHub soon.
Technical Details
For those interested in the implementation:
- Custom ESPHome component written in C++
- Uses ESP-IDF framework for optimal performance
- Learning matrix stored in flash with checksum validation
- Intelligent pump control with multiple trigger modes (manual, scheduled, water draw, etc.)
Questions? Want to Join?
Happy to answer any questions about the hardware design, software architecture, or installation process.
Interested in getting a PCB? Send me a message! I’m looking to build a small community of testers and contributors who can help refine the project and share their experiences with different installation scenarios.
The more people testing in different households, the better we can optimize the learning algorithms and features!