2.GATHERING PERTINENT INFORMATION - T-div-Engg-Exploration-2026-27/Repo-2 GitHub Wiki

AI-Based Smart Seed Sowing Machine

1.INTRODUTION

The AI-based seed sowing machine is a smart mechatronic system built to perform farming operations automatically with less human effort. It is capable of moving in the field, checking soil fertility and sowing seeds on its own .

The main aim of this project is to overcome the problems in traditional farming like fertilizer wastage and uneven sowing. It tries to develop an affordable and intelligent machine that can handle both fertilizing and sowing work.

The working of the system involves capturing soil data through camera and NPK sensors. This data is analysed by the ESP32 controller using AI to identify nutrient deficiency. Once analysed, the system controls the servo motor to drop the exact amount of required fertilizer and then places the seed at proper spacing. This helps in improving soil health, saving resources and increasing crop yield.

2.EXISTING SOLUTIONS TO THE PROBLEM

Existing solutions available for seed sowing and fertilizing:

  1. Manual Hand Sowing:Traditional,time-consuming,uneven.
  2. Hand-operated Rotary Seeder: Mechanically drops seeds at equal distance but no soil testing.
  3. Tractor-mounted Seed Drill: Used for large farms, high cost, no AI.
  4. John Deere ExactEmerge Planter: High-end commercial AI planter with sensors - very costly ~ 15 Lakhs.
  5. Automatic Seed Sowing Robot (Arduino based): Student DIY projects that only sow seeds, no fertility check.

Our project is different because we combine Soil Fertility Detection + Fertilizer Dosing + Sowing in one low-cost robot.

3.Table 1.1 Information Collected by various sources

Sl. No Components or parts used:Electronic Components or parts used:Mechanical Mechanism / Principle Identified Link
1 ESP32 + L298N 4WD Chassis Bluetooth controlled sowing robot https://www.instructables.com/SEEDS-SOWING-ROBOT/
2 Soil Moisture Sensor Seed Hopper flap Moisture based sowing https://www.engineersgarage.com/soil-moisture-detector/
3 ESP32-CAM - Image based NPK estimation using CNN https://en.wikipedia.org/wiki/Precision_agriculture
4 ESP32-CAM AI - Soil color to fertility correlation research https://farm.bot - FarmBot Open Source Farming Robot Website
5 Servo SG90 PVC Pipe Dispenser Servo controlled fertilizer drop https://www.instructables.com/Virtual-Presence-Robot/ - Has SG90 + L298N + ESP32-CAM
6 ESP32, Relay Drill mechanism Automated seed drilling https://www.youtube.com/watch?v=gLHz4E9SxLU
7 Arduino UNO, Ultrasonic Encoder Wheels Row spacing measurement https://learn.adafruit.com/adafruit-ultimate-gps/overview
8 Buck Converter LM2596 Plywood Frame Power distribution for farm robot https://www.instructables.com/Variable-Power-Supply-Buck-Converter/
9 ESP32 + Blynk - IoT soil data logging https://how2electronics.com/interface-capacitive-soil-moisture-sensor-arduino/
10 ESP32-CAM Plastic Bottle Hopper Low cost hopper design https://www.engineersgarage.com/soil-moisture-detector/
11 BO Motor 150 RPM Chain Drive Mobility in field https://www.youtube.com/watch?v=Rkj1DQQUjVo - Commercial Seed Sowing System
12 LCD 16x2 + ESP32-CAM - Displaying fertility report NPK % https://www.youtube.com/watch?v=sU87nVbK5aM - Arduino Seed Sowing
13 - Plough Tool Soil opening for seed https://www.youtube.com/watch?v=SOSxyH5nZxE - 3D Printed Seed Sowing Robot
14 Solar Panel 10W Solar Mount Solar powered agribot https://www.azorobotics.com/News.aspx?newsID=15867 - Solar Powered Robot 98.8% Accuracy
15 GPS Neo-6M - GPS based path following https://learn.adafruit.com/adafruit-ultimate-gps/overview
16 ESP32-CAM Soil Fertility AI Model SD Card Module Soil image dataset collection https://www.youtube.com/watch?v=Rkj1DQQUjVo
17 Servo MG996R Fertilizer Tank Precise dosing https://www.youtube.com/watch?v=sU87nVbK5aM
18 Relay 5V 4-channel Water Pump Irrigation after sowing https://how2electronics.com/automatic-plant-watering-system-using-arduino-uno/
19 IR sensor SEED Level Indicator Seed empty alert https://www.electronicwings.com/sensors/moisture-sensor
20 Battery 12v 7Ah Chasis Frame Portable Power https://robofactory.in/product/4wd-robot-chassis-kit/

4. Components & Sub-components Required (No NPK Sensor)

Component Type Working Principle / Construction / Interaction
ESP32-CAM Module Electronics - Main Controller + AI Sensor Single board does everything. OV2640 camera captures soil image, CNN model trained on Kaggle "Soil Fertility Dataset" (500+ images) estimates NPK% from color/texture. Also controls motors/servos via GPIO. Replaces costly Rs. 2800 NPK sensor.
Soil Moisture Capacitive Sensor Electronics - Support Sensor Gives moisture % (0-100%). Helps AI to confirm soil condition. Cost Rs. 120 only.
SG90 Servo Motors x 2 Electronics + Mechanical 180 deg, PWM controlled. Servo 1 opens fertilizer valve, Servo 2 opens seed valve. Plastic gears, sufficient for 1L bottle hopper.
L298N Motor Driver Electronics - Driver Dual H-bridge. Drives 2x BO motors using 5V logic from ESP32-CAM.
BO Motors 150 RPM Dual Shaft x 2 Mechanical - Actuator Low-cost DC motors with 65mm wheels. 2WD with front castor wheel is enough for prototype. High torque for soil.
Chassis - Recycled Plywood 50cm x 35cm Mechanical - Base Made from scrap plywood from college workshop. Light weight, strong enough for 3kg load.
Seed & Fertilizer Hopper - Plastic Bottles Mechanical - Storage Two 1L bottles cut into funnel shape. One for fertilizer, one for seeds. Connected to PVC pipe (1 inch) to guide material to ground.
Power - 3x 18650 Cells (Reused) + LM2596 Electronics - Power 3x 18650 cells reused from old power banks (0 cost). 12V output. LM2596 buck steps down to 5V 2A for ESP32-CAM and servos. Battery holder Rs. 100.

NPK is estimated by AI vision only.

5. Recent Information Check

All information is recent 2023-2026. We used latest method - ESP32-CAM with CNN for NPK estimation (published in IEEE 2023), not old 2015 chemical methods. All DIY links are from 2022-2025.

6. Direct Solutions Available with Images & Citation

1. Manual Broadcasting Seeder - Fig. 1

Hand rotated, low cost, no automation. Ref: http://balwaan.com

2. Tractor Mounted Seed Cum Fertilizer Drill - Fig. 2

Uses tractor, 9 rows, ~Rs. 90k, no soil testing. Ref: http://indiamart.com

3. John Deere ExactEmerge AI Planter - Fig. 3

AI planter, 99% spacing accuracy, ∼15 Lakhs. Ref: http://deere.com

4. Arduino Seed Sowing Robot (Instructables) - Fig. 4

Student project, fixed interval sowing only. Ref: http://instructables.com

5. Solar Powered Automatic Seeder - Fig. 5

Solar powered, eco-friendly but no fertility detection. Ref: http://youtube.com

7. Observations in Local Market, Online, College Labs - 5 Observations

Observation 1: Durgad Bail Agriculture Shop, Hubli Saw heavy iron seed drills needing bullocks. Heavy and costly. Idea: Use lightweight plastic bottle hopper to reduce weight and cost to under Rs. 200. [Fig. 6]

Observation 2: Online - http://Robu.in / Amazon ESP32-CAM Rs. 650, DC motor Rs. 150, SG90 Rs. 150. Confirms prototype under Rs. 5000 is possible.

Observation 3: Mechatronics Lab, College Observed L298N + BO motor + SG90 setup in line follower robot. Same exact mechanism can be reused for our fertilizer dispenser valve control.

Observation 4: Farmer Field Near Hubli-Dharwad Bypass Farmers throwing fertilizer uniformly with hands, wastage near bunds. Validates need for variable-rate dispensing using AI.

Observation 5: College Library + Workshop Found IEEE papers on "Soil Color to NPK Correlation using Deep Learning" and found scrap plywood, PVC pipes, 3D printer for hopper fabrication. Can make chassis in workshop saving Rs. 2000.

8. Cost Estimation

A. Existing Commercial Solution Cost Tractor Seed Drill = Rs. 85,000 + Tractor Rent Rs. 2,000/day = Too costly for small farmers.

B. Our Prototype - Low Cost Rs. 5000 Model

Sl. No Component Cost (Rs.)
1 ESP32-CAM Module (Controller + Camera) 650
2 DC Motors 150 RPM x 2 + 65mm Wheels x 2 + Castor Wheel 450
3 L298N Motor Driver Module 220
4 DC Motors x 2 for Fertilizer & Seed Dispenser 300
5 Capacitive Soil Moisture Sensor 120
6 LM2596 Buck Converter + Breadboard + Jumper Wires + Switch 350
7 Chassis - Plywood from Workshop Scrap + Screws + Fevicol 200
8 Hoppers*2+ 2ft PVC Pipe + Tape + Glue 150
9 Battery Pack - 3x 18650 used from Power Banks + Battery Holder (Rs.100) 100
10 Soldering, Nuts-Bolts, Poster Printing, Binding, Transport & Contingency 760
TOTAL COST - Rs. 3,300/-
Final Billing with 50% Buffer for Price Variation - Rs. 4,950/-

Budget Justification Paragraph : Soil fertility is detected using low-cost AI-based image processing through ESP32-CAM module. The CNN model is trained on Kaggle Soil Fertility Dataset (500+ soil images) to estimate N, P, K values from soil color and texture. A capacitive soil moisture sensor (Rs. 120) supports the AI decision. Mechanical parts like chassis and hoppers are fabricated using recycled plywood and plastic bottles from college workshop to minimize cost. Reused 18650 cells from old power banks are used for power supply. Hence total cost is restricted to Rs. 4950/- within given budget.