IoT-enabled Feeder System for Koi-Fish

A prototype highlighting more the use case of IoT to monitor the aquaculture inside the tank particularly Koi-Fish 🐟

Update: This is included in ICOM`24 IEEE conference! You may check the paper below for more details.

About the thumbnail: This is the ThingSpeak channel for IoT of our system!

OVERVIEW:

The Smart IoT-Based Feeder System for Koi Fish was developed to modernize aquaculture feeding and monitoring through Internet of Things (IoT) technology. Traditional aquaculture relies heavily on manual feeding and supervision, which can be time-consuming and inefficient. By integrating sensors, microcontrollers, and cloud-based monitoring, this system aims to automate the feeding process and maintain optimal water quality conditions for koi fish cultivation.

The system is built around a network of IoT sensors and actuators connected to microcontrollers and a cloud platform. It automates fish feeding cycles, monitors water temperature and level, and allows users to access real-time environmental data through a mobile application interface. This project represents a step toward more sustainable and data-driven aquaculture management.

Methodology

1.) System Architecture The feeder system integrates both hardware and software components connected through IoT protocols.
The main components are:

  • Microcontrollers: Raspberry Pi (RPi) and ESP32
  • Sensors: Water level sensor and DS18B20 temperature sensor
  • Actuator: SG90 servo motor for controlling food release
  • Connectivity: Wi-Fi network for cloud communication
  • Cloud Platform: ThingSpeak API for data storage and visualization
  • Power Source: Power bank as a portable power supply

The ESP32 collects sensor data (temperature and water level), processes it, and transmits it to the ThingSpeak cloud via Wi-Fi. The Raspberry Pi acts as a centralized power source, providing regulated power to all connected devices.

2.) Feeding Mechanism Design The automatic feeder mechanism uses a servo motor to open and close a feeding container that dispenses koi food pellets. The servo is triggered by a pushbutton signal, which emulates an automated feeding schedule. The motor’s motion log is recorded in the system as an indication of feeding activity.

The feeder enclosure was designed using corrugated cardboard, considering the space required for the main components—power bank, breadboard, sensors, and control units. Adjacent to the main housing is a pellet container that is mechanically linked to the servo motor. This structure allows consistent and controlled feeding cycles.

3.) Software Integration The firmware was programmed using the Arduino IDE for the ESP32 microcontroller. It handles:

  • Data acquisition from sensors
  • Transmission of data to ThingSpeak via HTTP requests
  • Control signals for the servo motor

The ThingSpeak API key authenticates the ESP32’s communication with the cloud. Sensor data is displayed in four fields on the ThingSpeak dashboard:

  1. Water temperature (°C)
  2. Water level (%)
  3. Motor actuation status
  4. Channel location

Additionally, a mobile GUI was developed using MIT App Inventor, enabling remote visualization of tank data and feeding history.

Video demonstration of temperature sensor functionality
Video demonstration of water level sensor functionality
Video demonstration of motor actuation functionality

FINDINGS:

1.) System Functionality The prototype successfully demonstrated IoT-based monitoring and feeding automation.
Key observations include:

  • The DS18B20 sensor recorded tank temperatures ranging between 33–37°C, accurately reflecting typical koi aquaculture conditions.
  • The water level sensor provided consistent readings, maintaining an 80% level near overflow during testing.
  • The servo motor operated reliably during feeding cycles, recording actuation events on the cloud dashboard.

Data from both sensors were continuously updated in real-time on the ThingSpeak channel, providing a clear visualization of environmental conditions.

2.) Graphical User Interface (GUI) The mobile application displayed:

  • Current temperature (e.g., 32.37°C)
  • Water level percentage
  • Motor actuation count

This provided an intuitive and mobile-friendly control panel for aquaculture managers, enabling remote monitoring without the need for physical inspection.

LEARNINGS:

The Smart IoT-Based Feeder System for Koi Fish successfully demonstrates how IoT technologies can be applied to modernize aquaculture practices. By combining microcontrollers, sensors, actuators, and cloud platforms, the system automates feeding, monitors water conditions, and provides accessible real-time data through mobile applications.

This project highlights the potential of IoT-driven solutions in achieving sustainable and intelligent aquaculture systems, reducing manual labor while improving fish welfare and production efficiency.

IEEE paper for our feeder system

HOW THIS PROJECT CAN BE FURTHER IMPROVED: Of course, I recognize that this is a rudimentary approach for this type of application where we just provided a simple prototype to deploy this into a water tank for one of our groupmates to use. A more robust system to check this in a dedicated app is suitable and also the enclosing for the sensor should be done in order to be more desirable for use. Aside from this other parameters for water quality was not checked but turbidity, salinity should be considered as this is a vital indicator of the environment and will affect the health of the koi fished.

Overall, the system to be operational in an industrial and commercialized settings will have to run more strict test and apparatus but this system is a stepping for many researchers on how to apply this in their MCUs!

PS. Shoutout to my groupmates jaimie and patrick for being involved in the project! Appreciate you guys~!