IoT-Based Autonomous Agribot: An Automated Seed Sowing System using ESP8266

We Built a Robot That Sows Seeds Using Wi-Fi — Here’s Everything You Need to Know

By Md. Ariful Islam Taiyeb, Shemin Hasan, Md. Abdulla Al Amin Shihab, Mahmudul Hasan & S.M Salman Malonchi Department of Software Engineering, Daffodil International University, Dhaka, Bangladesh Published as an IEEE Research Paper

The Problem: Why Does This Even Matter?

Picture a rice farmer in rural Bangladesh. Every season, he spends days crouching under the sun, pressing seeds into the soil one by one — row after row, hour after hour. It is exhausting, slow, and prone to human error. Seeds get spaced unevenly, planting depth varies, and fatigue sets in quickly.

Now multiply that challenge across millions of farmers and billions of acres of farmland. The world grows more food every year — but the workforce available to plant it is shrinking. Young people are moving to cities. Labor costs are rising. Climate change is making growing seasons unpredictable.

This is the crisis that modern agricultural automation is trying to solve. And our team of five software engineering students from Daffodil International University decided to contribute our part: a low-cost, Wi-Fi controlled Seed Sowing Robot.

What Is an IoT Agribot?

Before we dive into what we built, let’s break down the terminology for those who are new to this field.

IoT stands for Internet of Things. It refers to everyday physical devices — machines, sensors, vehicles — that are connected to the internet and can send or receive data. Your smart TV, fitness tracker, or even a smart thermostat — all IoT devices.

An Agribot is simply a robot designed to work in agriculture. Think of it as a machine that replaces manual labor in a field — planting, watering, harvesting, or monitoring crops.

Put them together, and an IoT Agribot is a farming robot that is connected to the internet, meaning it can be monitored and controlled remotely using a smartphone or computer. No need to be physically present next to the machine.

Our robot specifically focuses on seed sowing — the process of planting seeds in the ground. That is the first and one of the most critical steps in farming.

The Technology Behind the Robot

The Brain: ESP8266 (NodeMCU)

The central controller of our robot is a tiny microcontroller chip called the ESP8266, housed on a development board known as NodeMCU.

Think of it like this: if the robot were a human body, the ESP8266 would be the brain. It processes all incoming instructions and sends signals to the motors telling them what to do.

What makes the ESP8266 special is that it has built-in Wi-Fi capability. Most basic microcontrollers (like the popular Arduino Uno) need an external Wi-Fi module to connect to the internet. The ESP8266 does this natively — making it far more affordable and compact.

Key specs: • Works on 3.3V • Supports 802.11 b/g/n Wi-Fi • Has multiple GPIO (General Purpose Input/Output) pins for connecting motors and sensors • Can be programmed using Arduino IDE (a beginner-friendly coding environment)

Cost: Approximately 200–400 BDT (about $2–4 USD)

The Muscle: L298N Motor Driver

The ESP8266 is powerful at processing information — but it cannot directly power motors. Motors need much more electrical current than a microcontroller can safely provide.

This is where the L298N Motor Driver comes in. It acts as a bridge — receiving low-power control signals from the ESP8266 and using them to drive high-power current to the motors.

Think of it like a volume control knob on a speaker system. The knob (ESP8266) sends a small signal, but the amplifier (L298N) does the actual heavy lifting to produce the sound (motor movement).

The L298N can control: • Direction (forward or backward for each motor independently) • Speed (using PWM — Pulse Width Modulation signals) • Two motors simultaneously (we used two L298N units to control all four motors)

Pins used from ESP8266: • D0, D1, D2 → Control motor 1 and 2 • D3, D4, D5 → Control motor 3 and 4

The Legs: 4 Gear Motors + 4 Wheels

Our robot moves using four DC gear motors, each connected to a wheel. Gear motors are DC motors combined with a gearbox — this combination trades speed for torque (rotational force), which is exactly what you need for a robot carrying weight across a field.

By controlling which motors spin and in which direction, we can make the robot: • Move forward (all four motors forward) • Move backward (all four motors backward) • Turn left (right-side motors forward, left-side motors slower or reversed) • Turn right (left-side motors forward, right-side motors slower or reversed)

This is the same basic principle used by tanks, wheelchairs, and many small robots — called differential steering.

The Planter: Servo Motor

While the gear motors handle movement, the servo motor handles the seed dispensing mechanism.

A servo motor is different from a regular DC motor in a key way: it can rotate to a specific angle and hold that position precisely. This makes it perfect for controlled, repeatable movements — like opening and closing a seed gate.

In our robot, the servo motor is connected to a small container that holds seeds. When the robot is in position to plant, the ESP8266 sends a signal to the servo, which rotates and opens the gate, releasing a seed. Then it closes again. This happens repeatedly as the robot moves.

How it works step by step: Robot moves to a position ESP8266 sends PWM signal to servo Servo rotates to “open” position (e.g., 90 degrees) Seed falls into the soil Servo returns to “closed” position (0 degrees) Robot moves to next position Process repeats

The Power System: 3 × 3.7V Batteries + BMS

The entire robot runs on three 18650 lithium-ion batteries, each with a voltage of 3.7V. Connected in series (one after another), they produce a combined 11.1V — enough to power the motors and the ESP8266 simultaneously.

But raw battery packs can be dangerous. Overcharging, over-discharging, or short circuits can cause fires or permanent damage. That is why we added a 3S 20A Battery Management System (BMS).

The BMS acts as a safety guardian: • It balances the charge across all three cells • It cuts off power if voltage drops too low (preventing over-discharge) • It limits current to prevent dangerous overloads (rated for up to 20 Amperes) • It allows safe charging through a standard charging port

Total energy capacity: ~37–40 Wh — enough for about 1–2 hours of continuous operation depending on load.

The Remote: Blynk Mobile App

The robot is controlled entirely through a smartphone app called Blynk.

Blynk is an IoT platform that allows you to build custom dashboards and control panels for microcontroller-based projects — without writing any mobile app code. You set up the interface in the Blynk app, link it to your ESP8266, and control your hardware from anywhere in the world via the internet.

How our control interface works: • Up arrow button → Robot moves forward • Down arrow button → Robot moves backward • Left arrow button → Robot turns left • Right arrow button → Robot turns right • Seed button → Activates servo motor to dispense a seed • Stop button → All motors halt immediately

Each button in the app corresponds to a “virtual pin” in the Blynk system. When you press a button, Blynk sends a signal over the internet to your ESP8266 (which is connected to the same Wi-Fi network). The ESP8266 reads the signal and activates the correct pins on the L298N motor driver.

Latency: In our tests, the response time between pressing a button and the robot reacting was less than 200 milliseconds — essentially instant.

System Architecture: How It All Connects

Here is the full data flow from user action to robot movement:

All of this happens within milliseconds. The entire system requires only a Wi-Fi network — no wires connecting the phone to the robot, no line-of-sight requirement, no specialized hardware beyond a smartphone.

The Code: How We Programmed It

The robot was programmed using Arduino IDE — a free, beginner-friendly coding environment. The code is written in Embedded C, a variant of the C programming language optimized for microcontrollers.

Here is a simplified breakdown of what the code does:

Library Imports

Pin Definitions

Main Loop

Motor Control Functions (example)

Seed Dispensing

The full source code is publicly available and can be adapted for similar agricultural robotics projects.

Testing and Results

We tested the robot across four scenarios — one for each direction of movement — and measured: • Whether seeds were successfully dispensed • How long each operation took • Speed consistency

Performance Table

Key findings: • The robot successfully dispensed seeds in 100% of test cases • Left and right turns operated at lower speed (60%) to maintain stability during turns • Forward and backward movements used higher speed (70%) for field coverage efficiency • Wi-Fi response latency was consistently below 200ms • The servo mechanism was reliable across all tests — no seed jams or misfires

Observed limitation: Performance on significantly uneven terrain (like a rocky field) was reduced. The four-wheel design without suspension had some stability issues on bumpy surfaces. This is a known limitation and a target for future improvement.

How This Compares to Existing Solutions

To put our work in context, here is how it compares to other approaches in the field:

The key differentiator is cost and accessibility. Industrial agricultural robots exist — but they cost thousands to tens of thousands of dollars. Our robot can be replicated for under $40 in components, using freely available software tools.

What’s Coming Next: Future Work

We are already planning the next iteration of this system. Here is what we want to build:

Soil Moisture Sensor Integration Add sensors that detect how wet or dry the soil is before planting. The robot would only dispense seeds when moisture levels are within the ideal range for germination — saving seeds and improving crop success rates.

Autonomous Navigation with GPS Replace manual control with automatic path planning. Using a GPS module and field boundary data, the robot would calculate optimal rows, navigate autonomously, and cover the entire field without human input.

AI-Based Smart Farming Train a machine learning model to analyze soil conditions, weather data, and crop type — and automatically decide the optimal planting density, depth, and timing. The robot would become truly intelligent, not just remote-controlled.

Multi-Robot Coordination Deploy a fleet of robots that communicate with each other to divide a large field and work in parallel, significantly reducing total planting time.

What We Learned

Building this robot taught us far more than any textbook could. A few key takeaways:

Hardware is humbling. Code either compiles or it doesn’t. Hardware either works or it doesn’t — and diagnosing why it doesn’t is a completely different skill. We spent more time troubleshooting wiring and motor driver configurations than writing code.

IoT makes everything more powerful. The moment we got the robot responding to our phone over Wi-Fi, the entire project felt real in a way it hadn’t before. Remote control transforms a machine from a demo into a tool.

Constraints drive creativity. Our entire budget was limited. That forced us to choose components carefully, optimize our code, and find clever solutions to problems that a bigger budget might have papered over.

Start simple, then iterate. We did not start by trying to build a fully autonomous AI-powered robot. We started with: can we make four wheels move? Then: can we add seed dispensing? Then: can we control it remotely? Each small win built the foundation for the next.

Conclusion

Agriculture needs innovation. Bangladesh — a country where farming employs over 40% of the workforce — needs affordable innovation. Our team built a small but meaningful proof of concept: a seed sowing robot that can be built for under $40, controlled from a smartphone, and deployed without any special technical expertise.

We are not claiming to have solved agricultural automation. But we have shown that the tools are accessible, the technology is ready, and the impact is real.

If you are a researcher, maker, farmer, or student interested in agricultural robotics — the code is open, the components are cheap, and the problems are vast. Let’s build something.

Connect with the team: • Md. Ariful Islam Taiyeb | Daffodil International University • Shemin Hasan | Daffodil International University • Md. Abdulla Al Amin Shihab | Daffodil International University • Mahmudul Hasan | Daffodil International University • S.M Salman Malonchi | Daffodil International University

Full source code available at:

https://docs.google.com/document/d/e/2PACX-1vTiAZRaWYc5d6cuiQ0CTYPoAK2IaQZfLq3hshcZNGy-fu3zjyOf4HfFH5frXS6Uqx3ig1hI1tO4hFwV/pub

This article is based on the IEEE conference paper: “IoT-Based Autonomous Agribot: An Automated Seed Sowing System using ESP8266”

Tags: #IoT #SmartAgriculture #ESP8266 #NodeMCU #Robotics #AgriTech #Bangladesh #EmbeddedSystems #OpenSource #Innovation #SoftwareEngineering #DaffodilUniversity