Department of Electronics and Communication Engineering

Netaji Subhas University of Technology

B. TECH PROJECT-I (ECECC-22)

Smart Farming System using Machine

Learning
Under the guidance of
Prof. Satya Prakash Singh
Submitted By
PRITIKA 2020UEC2537
ANANYA PRATAP SINGH 2020UEC2505
SHASHWAT 2020UEC2535
Session 2023-24 YASH CHAURASIA 2020UEC2534
Semester: 07 (Mid-Sem Review)
 INTRODUCTION
Smart Farming system is an IoT and ML-powered agriculture assistant that helps farmers
improve their yields and reduce their costs. It collects data from sensors in the field, such as
soil moisture, temperature, and humidity, and uses ML algorithms to provide farmers with
personalized recommendations for irrigation, fertilization, and pest control.

ADVANTAGES
Increased crop yields: help farmers optimize their crop management practices.
Reduced costs: helps farmers to reduce their input costs by providing them with precise
recommendations for irrigation, fertilization, and pest control.
Improved sustainability: reduce their environmental impact by helping them to use water and
other resources more efficiently.
Enhanced decision-making: Smart Farm Advisor provides farmers with real-time data and insights
that they can use to make better decisions about their crops and their business.
 OBJECTIVE
1. Develop an agriculture system that uses sensors to monitor soil moisture,
temperature, NPK levels, and pH.
2. Use machine learning to develop predictive models that can help farmers make
informed decisions about crop selection, planting, harvesting, and irrigation.
3. Help farmers improve their crop yields and profitability by providing them with
personalized recommendations for crop selection, crop rotation, and irrigation.
4. Weather forecasting: A weather forecasting tool could provide farmers with up-
to-date information on current and upcoming weather conditions. This could help
farmers plan their tasks and make decisions about when to plant, water, and
harvest their crops.
5. Market prices: The app could include a feature for tracking current market prices
for various crops. This could help farmers make informed decisions about when to
sell their crops and at what price.
LITERATURE SURVEY
 GAPS IDENTIFIED
Farmers depend on labs and have less real-time awareness of soil conditions: Farmers typically
rely on lab tests to determine the nutrient levels and pH of their soil. This can be expensive and
time-consuming, and it does not provide farmers with real-time information about their soil
conditions.
Farmers lack access to data and predictive analytics: Farmers often do not have access to the
data and predictive analytics tools that they need to make informed decisions about irrigation and
crop selection. This can lead to suboptimal irrigation practices and crop selection decisions

Lack of access to market information and pricing: One of the major challenges that farmers face
is a lack of access to timely and accurate information about market prices for their crops. Without
this information, it is difficult for farmers to make informed decisions about when to sell their crops
and at what price.
Lack of information about weather conditions: Another challenge that farmers face is a lack of
access to accurate and timely information about current and future weather conditions. This can
include information about temperature, precipitation, and other factors that can affect crop growth
and development

Goal:
The aim of this project is to develop a Smart Farming system using machine learning to
help farmers improve their crop yields and profitability, by providing real time update.
 METHODOLOGY ADOPTED
Literature Selection of Selection of
review objective parameters

Deploying ML Design of
Data collection
algos interface/app

Optimisations Prediction and

conclusions
 HARDWARE

Temperature and Soil Moisture NPK Sensor

Humidity Sensor (DHT11) Sensor (FC-28)
DHT11 is a low-cost digital FC-28 soil moisture sensor is a soil The soil NPK sensor is suitable NodeMCU ESP8266 Pump (12V DC)

sensor for sensing hygrometric transducer that can read for detecting the content of The NodeMCU is a development Pumps water to and from the

the amount of moisture present in the N, P, and K. board that is based on the plant. It is controlled via relay
temperature and humidity.
ESP8266 microcontroller. It has a which is controllable via
soil surrounding it.
built-in web server, which allows ESP32
you to connect to the NodeMCU
ESP 32 using a web browser.
Micro-controller which
can be programmed with
Arduino IDE. All the
sensors and actuators are Relay
connected to esp32 to
their respective pins and The relay is a programmable switch
these pins are specified that can be used to control
accordingly in the ON/OFF electrical devices. The
program for respective relay can be programmatically
sensors controlled by ESP32
Fig. Smart Farming System
DATASET USED
 ML ALGORITHMS
SVM is a powerful supervised algorithm that works best on smaller
datasets but on complex ones. Support Vector Machine, abbreviated
as SVM can be used for both regression and classification tasks, but
CROP SUGGESTION generally, they work best in classification problems
KNN has low accuracy as compared to SVM.

KNN Model can be used YIELD PREDICTION

ML algorithms, K-nearest neighbor, support vector regression

(SVR), Naive Bayes, etc. can be used to predict soil dryness based
SOIL MANAGMENT
on precipitation and evaporative hydrology data.
APP DEVELOPMENT FLUTTER
Crop Suggestion

display sensor real-

time data Google
Firebase
Weather Forecast

Market Price

Cloudflare
Fertilizer suggestion
WORK DONE TILL NOW
Literature survey has been done, where we have studied papers in the domain of inculcating machine
learning in agriculture.

After realising the gaps in the progress in the realm of application of smart agriculture, we
have identified the objective of developing a agriculture system that uses sensors to
monitor soil characteristics and deploying the machine learning model.

Parameters selected for study are moisture, temperature, NPK level.

Suitable sensors are selected to study these parameters.

Methodology to be adopted in implementing the project has been decided.

 FUTURE WORK TO BE DONE
Hardware connections and web/app development:
1. Develop more efficient and cost-effective hardware for collecting data about the environment
and crop health.
2. Develop more user-friendly web and mobile/web apps for farmers to interact with the smart
irrigation system.
3. Explore the use of artificial intelligence (AI) to automate tasks such as irrigation scheduling and
pest and disease management.

Machine learning on agriculture data set:

4. Develop new machine learning models that can predict crop irrigation requirements more
accurately.
5. Train machine learning models on larger and more diverse datasets to improve their
generalization ability.
6. Explore the use of machine learning to predict other aspects of crop production, such as crop
yields and pest and disease outbreaks.
 EXPECTED OUTCOMES
We can make the decision-making for the farmers easy using this project.

We can be able to predict suitable crops that can be grown in a

particular area.

Higher crop productivity.

Decreased use of water, fertilizer, and pesticides, which in turn

keeps food prices down and reduces food crisis.

Reduced impact on natural ecosystems and maintains Soil Health.

Increased crop sustainability.

 REFERENCES
1. A.Anitha, Nithya Sampath, M.Asha Jerlin (Year-2020): Smart irrigation using IOT
2. Pedro Corista, Diogo Ferreira, Joao Giao, Joao Sarraipa, Ricardo Jardim
Goncalves:IOT agriculture system using FIWARE
3. G. Sushanth, S. Sujatha: IOT Based Smart Agriculture system
4. Nermeen Gamal Rezk, Ezz El-Din Hemdan, Abdel-Fattah Attia, Ayman El-Sayed,
Mohamed A. El-Rashidy: IoT based smart farming system using machine learning
algorithms
5. M.W.P Maduranga, Ruvan Abeysekera 2020: Machine learning applications in IOT-
Based agriculture and smart farming: a review
6. K. Jyostsna Vanaja, Aala Suresh, S. Srilatha, K. Vijay Kumar, M. Bharath: IOT
based Agriculture System Using NodeMCU
 THANK YOU
PRITIKA 2020UEC2537
ANANYA PRATAP SINGH 2020UEC2505
SHASHWAT 2020UEC2535
YASH CHAURASIA 2020UEC2534