Automated Pool Fill and Monitoring System

This project describes a DIY engineer’s journey to replace a failing pool system with a fully automated, sensor‑driven solution. Using a Raspberry Pi, low‑power Arduinos, and open‑source software, the system continuously monitors water level, temperature, pH, ORP, filter pressure, electrical usage, and pump operation, refilling the pool only when necessary.

Story

In the summer of 2015, my sons and I tackled a cascade of problems in our home pool—faulty water lines, leaky valves, a subpar filter, and an ineffective pool company. Rebuilding the entire system became a mission: clean the pool, install new components, and, most importantly, give it intelligent automation.

The Idea

The core goal was to replace the old mechanical setup with a network of sensors and a control hub that could run the pool autonomously. We chose the Autopilot management system for salt‑water conversion and basic chemical dosing, while the bulk of the automation was handled by custom hardware.

Getting Started

After upgrading the filter, pump, and valves, we introduced a Raspberry Pi‑based system to read sensor data and drive the water fill valve. The Pi communicates with battery‑powered microcontrollers that handle level, temperature, and pressure measurements.

Automation – The Beginning

We needed a way to detect water level and trigger the fill valve. Initially, a Milone liquid‑tape resistance sensor was selected, but a water ingress failure prompted a switch to a low‑cost float‑switch solution. A low‑power Moteino‑R5 reads the float status and sends the data over an RFM69 transceiver to the Pi.

Filling the Pool

Water is supplied via the existing irrigation system. A 24 V AC sprinkler valve, controlled by a 240 V to 24 V transformer and an optically isolated solid‑state relay, is energized only when the pool is below the target level and the sprinklers are idle. The Rachio API is queried every minute to confirm that no lawn zone is using water.

Power Management

Pump operation is monitored through a Brultech GEM electrical monitor. When the pump draws more than 5 kW, the system suspends filling to avoid back‑pressure loss. A MightyHat UPS provides two‑hour backup for the Pi, and a manual “Stop Fill” button shuts down the transformer if needed.

Monitoring Sensor Suite

• Water level (float switch)
• Pool temperature (RTD probe)
• Filter pressure (100 PSI sensor)
• pH and ORP (Atlas Scientific sensors in a flow cell)
• Electrical load (GEM, 5 kW max)
• Sprinkler status (Rachio API)
• Acid tank level (DFRobot liquid‑level sensor)

Water Level Sensor – New Method

The float‑switch setup uses two 1/4″ float switches mounted inside a weather‑sealed enclosure. The upper and lower floats report open/closed states; the system interprets the combination as low, midway, or full. The sensor node sends a 0, 1, or 2 to the Pi every minute.

Temperature Sensor – Floating Enclosure

A temperature probe is placed in a chlorine floater that rests on the pool surface, ensuring reliable readings and protecting the sensor from accidental damage.

Acid Level Monitoring

A DFRobot liquid‑level sensor on the acid tank triggers a notification when the tank is low, allowing manual refilling before the Autopilot can add acid.

Web Interface

A Flask‑based dashboard displays real‑time gauges, system status LEDs, battery health, and controls for manual fill, pump programs, and notification settings. The interface also hosts an Alexa skill via Flask‑Ask, letting users query pool status or start/stop filling with voice commands.

Historical Graphing

Sensor data is logged to a MySQL database and forwarded to InfluxDB; Grafana visualizes trends in temperature, pH, ORP, and water usage. The dashboards help spot anomalies and optimize pool chemistry.

Notifications

The system supports log files, console debug, Pushbullet, email, and Twilio SMS. Users can customize which events trigger which channel directly from the web panel.

Alexa Integration

A custom Alexa skill, built with Flask‑Ask, provides spoken updates on temperature, pH, pump status, and can command the system to start or stop filling.

Conclusion

Version 3.5.0 of the codebase modularizes the logic into reusable Python functions, improving maintainability and reliability. The project demonstrates that a homeowner can replace a legacy pool system with a low‑cost, open‑source, and highly reliable automation stack, delivering real‑time insights and hands‑free control.

Source: Pool Fill Control