SRM VALLIAMMAI ENGINEERING COLLEGE

(An Autonomous Institution)

SRM Nagar, Kattankulathur – 603 203

DEPARTMENT OF INFORMATION TECHNOLOGY

VIII SEMESTER

1908808 – PROJECT (PHASE-II)

FIRST REVIEW ( DATE : 11 /01/2024)

SMART AGRICULTURE USING MACHINE LEARNING AND AI

Team Members:
Under the guidance of
Pavithra.S 142220205058
Mrs. S.Shenbagavadivu, Preethi.R 142220205063
Assistant Professor, Priya dharshini.K 142220205065
Department of Information Technology Sakithya.P 142220205073
 INTRODUCTION

• The aim of this work is to help an individual cultivate crops efficiently and hence achieve high productivity at low cost. It

also helps to predict the total cost needed for cultivation. This would help an individual to pre-plan the activities before

cultivation resulting in an integrated solution in farming.

• In second phase of our project we going to create the model for crop recommendation, pest and disease detection system,

irrigation notification system using ML and AI algorithms and going to integrate with our web app that is designed and

implemented in previous phase.

• we studied and compared the efficiency of different ML algorithms such as decision tree, Gaussian Naïve Bayes, Support

vector machine(svm), Logistic Regression, Random forest and XGBoost algorithms.

• With our study we observed that the accuracy of Naïve Bayes algorithm and XGBoost is high compared to other algorithms

so we created prediction model using Naïve bayes and XGBoost algorithms.

 OBJECTIVE

• Optimize farming processes.

• Smart farming systems do not require much prior expert knowledge in order to start your
own cultivation

• Provide the farmer with an added value in the form of better decision-making or more
efficient operation and management of his farm.
 LITERATURE SURVEY

S.No Title Author Year Methodology Inference Merits Demerits

1 Towards Shaikh, T. A., 2022 Support Scalable and Does not Supervised
leveraging the Rasool, T., & Vector parallelizable provide learning for
role of machine Lone, F. R. Machines algorithm for insights into classification
learning and (SVM) distributed the model's and
artificial computing. learning regression.
intelligence in curve.
precision
agriculture and
smart farming

2 Vertical farming Siregar, R. R. 2022 AlexNet Addresses No Deep learning

perspectives in A., Seminar, transfer consideration that
support of K. B., learning for model popularized
precision Wahjuni, S., through fairness- convolutional
agriculture using & Santosa, E. domain aware neural
artificial adaptation. optimization. networks.
intelligence: A
review
 LITERATURE SURVEY

S.No Title Author Year Methodolog Inference Merits Demerits

y

3 IoT-equipped and Qazi, S., 2022 Locally Handles No insights Manifold

AI-enabled next Khawaja, B. Linear continuous into feature learning for
generation smart A., & Embedding and discrete importance preserving
agriculture: A Farooq, Q. (LLE) data in a in the model. local
critical review, U. unified neighborhoo
current challenges framework. d
and future trends relationships.

4 Machine learning Shaikh, T. A., 2022 Self- Evaluates Lacks Incremental

for smart Mir, W. A., Organizing model theoretical learning for
agriculture and Rasool, T., & Incremental uncertainty analysis for unsupervised
precision farming: Sofi, S. Neural through the proposed and online
towards making Network Bayesian algorithm. learning.
the fields talk (SOINN) neural
networks.
 LITERATURE SURVEY

S.No Title Author Year Methodolog Inference Merits Demerits

y

5 Enabling Sharma, A., 2022 Q-Learning Addresses Does not Reinforceme

smart Georgi, M., with Function missing data address nt learning
agriculture by Tregubenko, Approximati imputation model's using
implementing M., Tselykh, on (QFA) with multiple robustness function
artificial A., & imputation. against approximatio
intelligence Tselykh, A. outliers. n.
and
embedded
sensing
 DRAWBACKS OF EXISTING WORK

• Static Predictions: Existing systems rely on outdated algorithms, leading to less accurate yield forecasts
and potential losses.

• Limited Mapping Capabilities: Without real-time geo-mapping, farmers often face challenges in accurate
field delineation and land management.

• Manual Irrigation Control: Lacking adaptive irrigation features, farmers manually adjust water
distribution, leading to potential water wastage.

• Reactive Pest Management: Current platforms lack predictive capabilities, causing farmers to react to
pest infestations rather than preventing them.

• Rigid User Interface: The static and non-customizable UI of existing systems doesn't cater to the diverse
needs of modern farmers.
 DRAWBACKS OF EXISTING WORK

• Local Data Storage: Limited to local storage, farmers can't access their data remotely, hindering efficient farm
management.

• Basic Weather Tracking: While they might provide current weather updates, existing systems lack in-depth weather
analytics, leaving farmers with limited insights.

• Inefficient Resource Utilization: Without AI-driven recommendations, there's often overuse or underuse of fertilizers
and pesticides, impacting both costs and crop health.

• Fixed Architecture: The monolithic design of many existing platforms makes it challenging to integrate new features
or technologies, limiting future growth.

• Solo Farming Focus: Most current systems don't support collaborative features, making it challenging for teams or
communities of farmers to share insights and collaborate.
 PROPOSED SYSTEM

• AI-Driven Predictions: Leveraging advanced machine learning algorithms, the system offers superior
yield predictions, optimizing farm productivity.
.

• Real-Time Geo-Mapping: With high-precision GPS integration, the platform provides real-time field
mapping, ensuring accurate land management.

• Adaptive Irrigation: The system's smart irrigation module adjusts water distribution based on real-time
soil moisture and weather forecasts, promoting water conservation.

• Integrated Pest Management: Using AI analytics, the platform predicts potential pest infestations,
allowing for proactive measures and reducing crop damage

• Dynamic UI Customization: The adaptive UI, with dynamic background customization, offers a
personalized user experience, enhancing user engagement.
 PROPOSED SYSTEM

• Predictive Crop Recommendation: A product recommendation system is a machine learning application with the
. cropping suggestions for users might like to engage with.

• Advanced Weather Analytics: The system not only tracks real-time weather but also analyzes historical weather patterns,
providing comprehensive insights for crop management.

• Resource Optimization Algorithms: AI-driven algorithms ensure optimal usage of fertilizers, pesticides, and other
resources, reducing costs and environmental impact.

• Scalable Architecture: Designed with a modular approach, the platform can easily integrate future technological
advancements, ensuring it remains at the forefront of agri tech solutions.

• Collaborative Farming Support: The system supports multi-user collaboration, allowing teams of farmers or agronomists
to work together, share insights, and make collective decisions.
ARCHITECTURE DIAGRAM
CROP RECOMMENDATIONARCHITECTURE DIAGRAM
 METHODOLOGY

PREDICTIVE CROP RECOMMENDATION SYSTEM:

1) NAÏVE BAYES: Random Forest classifier was found huge ability to predict crop yield. From different outputs, it
shows that Random Forest is an efficient learning algorithm to analyze crop at current climatic condition and has a huge
exactness in data investigation.

2) RANDOM FOREST: Naive Bayes, a supervised learning algorithm puts forth in the way to achieve it. The seed data
of the crops are collected here, with the appropriate parameters like temperature, humidity and moisture content, which
helps the crops to achieve a successful growth. In addition as the software, a mobile application for Android is being
developed.

3) XGBOOST:XG Boost model gives better performance than all other models and give high accuracy in prediction
based on the parameters needed for cropping. In future, it is planned to optimize the hyper parameters of each traditional
algorithms and should be tested with new data sets.
 PROPOSED MODULES

• Predictive cropping recommendation

• Pest and Disease Management

• Weather Tracking

• Irrigation Scheduling and Monitoring

 PROPOSED MODULES DESCRIPTION

• Predictive Cropping recommendation: This work is to help an individual cultivate crops efficiently and hence achieve
high productivity at low cost. It also helps to predict the total cost needed for cultivation. This would help an individual
to pre-plan the activities before cultivation resulting in an integrated solution in farming.

• Pest and Disease Management: Receive timely alerts and recommendations for pest and disease control. Access
integrated databases to identify and address issues promptly, safeguarding your crops.

• Weather Tracking: Stay informed about changing weather conditions with real-time updates and forecasts. Utilize
historical weather data to make informed decisions for your farming activities.

• Irrigation Scheduling and Monitoring: Optimize water usage and enhance irrigation practices. Set up customized
schedules based on crop needs and monitor water usage to promote sustainable farming.
 WORK PLAN FOR REMAINING WORK

Pest and Disease Management: Receive timely alerts and recommendations for pest and disease control. Access
integrated databases to identify and address issues promptly, safeguarding your crops.

Weather Tracking: Stay informed about changing weather conditions with real-time updates and forecasts. Utilize
historical weather data to make informed decisions for your farming activities.

Irrigation Scheduling and Monitoring: Optimize water usage and enhance irrigation practices. Set up customized
schedules based on crop needs and monitor water usage to promote sustainable farming.
HARDWARE REQUIREMENTS

• Desktop or Laptop computer with Intel Core i5 or i7 processor capable of running python 3.11.1 and above.

• With minimum of 8 GB RAM Processor and above for performing Machine learning algorithms.

• And with minimum of 100 GB HDD(Hard Disk Drive) and above for storing and processing huge datasets .

SOFTWARE REQUIREMENTS

• Operating System that supports python platforms and capable for running Machine Learning algorithms such
as windows 10 and 11.

• With Python platforms such as Jupyter notebook downloaded using Anaconda navigator and capable of
performing ML algorithms with huge datasets.

• PyCharm platform with latest version, capable of performing python 3.11.1 version such as PyCharm
2022.2.3.

• Gooogle CoLab with Python 3 kernel CoLab notebook .

 REFERENCES

[1] Shaikh, T. A., Rasool, T., & Lone, F. R. Towards leveraging the role of machine learning and artificial intelligence in precision
agriculture and smart farming Computers and Electronics in Agriculture, 198, 107119, 2022.,

[2] Siregar, R. R. A., Seminar, K. B., Wahjuni, S., & Santosa, E. Vertical farming perspectives in support of precision agriculture using
artificial intelligence A review. Computers, 11(9), 135, 2022.,

[3] Qazi, S., Khawaja, B. A., & Farooq, Q. U Machine learning for smart agriculture and precision farming: towards making the fields talk”
Archives of Computational Methods in Engineering, 29(7), 4557-4597,2022.,

[4] Shaikh, T. A., Mir, W. A., Rasool, T., & Sofi, S Enabling smart agriculture by implementing artificial intelligence and embedded sensing”
Computers & Industrial Engineering, 165, 107936,2022.,

[5] Sharma, A., Georgi, M., Tregubenko, M., Tselykh, A., & Tselykh, A. Everything you wanted to know about smart agriculture” arXiv
preprint arXiv:2201.04754,2022.,

[6] Krishnan et al, Climate change effects on Western Ghats ecosystems. DST Report, 2021.,

[7] Kaur and Kaur, Smart precision agriculture using IoT. International Conference on Communication and Signal Processing (ICCSP), 2022.,

[8] Toraskar et al.IoT-based smart agriculture. IEEE International Conference on Electronics and Sustainable Communication Systems
(ICESC), 2021.,

[9] ICRISAT, Unlocking the potential of AI for smallholder farmers. IEEE Conference Publication, 2021.