Pond Monitoring Project

Scripts (mostly Python, so far) to monitor various aspects of a backyard koi pond - environmental conditions, including temperature, water level, and potential threats (e.g., predators). Nothing terribly fancy or intelligent, just some messing around with a mixture of Raspberry Pi, sensor devices, Hubitat, IP cameras and computer vision.

Project Overview

The project consists of several components:

• Temperature Monitoring:
  The temperature/ subfolder contains two scripts:

  • Hubitat Temperature Monitor:
    Polls a water temperature sensor, calculates current, high, and low temperatures over a rolling 24-hour window, and sends the data to a Hubitat home automation hub via API calls.
  • Cacti Temperature Charting Script:
    Reads air and water temperatures from 1-Wire sensors and prints the values (formatted for charting via Cacti or a similar platform).

• Computer Vision:
  (Coming soon) Scripts that use machine learning and computer vision to identity & alert for potential predators (focussing on Herons for now): training models/weights, analyzing video feeds, testing tools, etc.

• IP Camera:
  Some simple scripts written to work with consumer IP cameras to trigger Hubitat API calls when motion is detected.

  The ip-camera/ subfolder contains two scripts:

  • Dummy TCP Server:
    A simple Python script, intended to be run as a service, that creates a basic TCP listener. Useful for consumer IP cameras who can connect to other devices (such as SMTP or FTP) when motion is detected. This works in conjunction with:
  • Hubitat Motion Trigger A bash script, also intended to be run as a service, that monitors when SYN packets are sent to a specific port (via tcpdump), and then sends an API call to a Hubitat virtual device (such as a Virtual Switch). Can manage multiple source devices and Hubitat virtual devices from a single service by MAC address from a single instance.

• Water Level Monitoring:
  (Planned) Future scripts to monitor the water level in the pond.