ESP32 Sensor Devices in Home Assistant

Lately, I’ve been experimenting with ESP32 boards and ESPHome to bring smarter environmental control into an insulated storage area in our garage. This post walks through how I set up temperature and humidity sensors, automated a dehumidifier, and visualized everything with Grafana, all using Home Assistant.

Hardware & Setup

I’m using an ESP32 microcontroller paired with a DHT22 sensor to monitor temperature and humidity. The ESP32 runs ESPHome, which makes it easy to define sensor behavior in YAML and integrate directly with Home Assistant over Wi-Fi. Here’s the basic ESPHome config for the sensor:

sensor:
  - platform: dht
    pin: GPIO4
    temperature:
      name: "Storage Temperature"
    humidity:
      name: "Storage Humidity"
    model: DHT22
    update_interval: 30s

Once flashed and connected, the ESP32 reports sensor data to Home Assistant automatically.

Automation: Humidity-Triggered Dehumidifier

To act on the data, I set up an automation that turns on a smart plug when humidity rises above 60%. The plug powers a “dumb” dehumidifier, and turns off when humidity drops below the threshold. Here’s the Home Assistant automation:

alias: "Dehumidifier Control"
trigger:
  - platform: numeric_state
    entity_id: sensor.living_room_humidity
    above: 60
action:
  - service: switch.turn_on
    target:
      entity_id: switch.dehumidifier_plug
mode: single

I use a second automation to turn it off when humidity falls below.

Visualization with Grafana

To monitor trends over time, I send sensor data to InfluxDB and visualize it in Grafana. This lets me track humidity levels, temperature changes, and dehumidifier activity on a single dashboard. Grafana helps me:

• Spot seasonal humidity patterns
• Adjust automation thresholds
• Validate that the system is working as expected

Why This Setup?

This project gives me:

• Environmental awareness: I know when the air gets muggy before it damages our stuff.
• Energy efficiency: The dehumidifier runs only when needed.
• Data-driven control: I can tweak behavior based on actual trends.

It’s a simple setup, but it’s reliable, extensible, and satisfying to build.