SMART INDIA HACKATHON 2025

Team

❑ Problem Statement ID :-SIH25010

❑ Problem Statement Title:- Smart Crop Advisory System for
Small and Marginal Farmers
❑ Theme:- Agriculture, Food-Tech & Rural Development
❑ PS Category:- Software
❑ Team ID-
❑ Team Name:- krishnova
 Idea title
❑ Proposed solution ❑ The solution addresses the problem in the
➢ Developing a multilingual crop advisor app based on AI & ML, which following way:
will act as a companion to a farmer, right from planting the crop to • It increases farm yield, consequently raising
selling the crop. farmers’ profit.
• Recommend traditional crops as well as crops that can be grown in • It cut input cost.
that region based on location, weather conditions and soil texture( • Promote sustainable practices.
data from Soil & Land use Survey of India). • Strengthen food security.
• Based on the uploaded soil report, it will recommend how much and • Limit unreliable advice.
what type of fertilizer to use. • Provide technical support.
• Time to time recommend when to irrigate based on the moisture
present in the soil detected via a special sensor provided to the ❑ Innovation and uniqueness of the solution
farmer. • AI-driven Recommendations: Beyond
• Pest/disease detection via Deep Learning (CNN Model Training & weather alerts, our system gives actionable
Prediction) and OpenCV (For Image Processing) • Offline-first Approach: App works in offline
• Market price tracking using APIs provided by platforms like the Open mode with cached advisories.
Government Data (OGD) Platform India. • Localized Advisory System: Unlike generic
• Voice support for low-literate users. weather apps, our solution provides.
• Feedback and usage data collection for continuous improvement. • Sustainability-Oriented: The solution saves
water, reduces costs, and improves yield.
2
 TECHNICAL APPROACH
Technologies to be used Methodology and process for implementation
❑ We use different technologies such as different programming
Registration/login
languages, Backend & server-side, Data Base, AI/ML & Analytics, • Enter mobile no.->OTP Verification-
cloud deployment, Hardware & IOT and other integrations, details •
>Set Password
Enable permissions(Location, Camera,
given below: Python, [Link], or java Microphone, Notification.
Farmer profile setup
• Enter Name, Address, Pin Code.
• Enter Land Size ->Upload/Scan Soil
Language Report
Data Processing and Analysis
• Location Mapping(GPS/Pin code)
Backend Framework My SQL / Postgre SQL(farmer profile, • Soil Data + Govt. Agri Database
soil data) • Weather API integration
Crop recommendation Engine
Relational DB • Suggest best crop(traditional+alternative)
MangoDB / firebase(real time updates and
• AL$ML based ranking.
Database notification
NoSQL DB Advisory & learning resources
• Crop Tuitorial(video/audio/pdf in local
Python Tensor Flow /
Farmer Application Libraries PyTorch(disease
•
lang.)
Step-by-step cultivation guidance
detection model
AI/ML & • Fertilizer and irrigation advisory
Analytics Open CV Image processing for leaf
disease Disease detection & Alerts
• Farmer scan leaf image(camera)
Scikit-learn Crop Recommendation model • AI model detects Disease and provides a
cure/solution.
Platforms
Cloud and Notification and continuous support
deployment • Alerts on weather, pest outbreaks
Flutter kotlin Firebase
• Market price updates
Docker & For scalable deployment
kabernetes
3
IoT & Soil moisture sensor
/weather station APIs
hardware
 FEASIBILITY AND VIABILITY
Feasibility Analysis Potential Challenges and Risks
• Digital literacy: Many marginal farmers may struggle
❑ Technical Feasibility
with mobile apps.
• Uses readily available technologies: mobile apps
• Internet connectivity: Rural areas often have poor data
(Android), cloud APIs, government open datasets (Soil
access.
Health Card, IMD weather, [Link]).
• Data quality issues: Incomplete or outdated soil reports
• Core AI/ML modules (crop recommendation, disease
/ weather data.
detection) are implementable with existing open
• Adoption resistance: Farmers may hesitate to shift from
datasets and lightweight ML models.
traditional crops.
• IVR/SMS fallback ensures inclusivity for farmers
• Language & usability: The Complexity of app UI may
without smartphones.
hinder adoption.

❑ Operational Feasibility ❑ Strategies for Overcoming These Challenges

• Can be deployed at the district/block level with farmer • Digital literacy:- Provide tutorial videos + voice
groups (FPOs, cooperatives). guidance in Punjabi.
• Integration with existing government databases makes • Internet connectivity:-Build offline-first app; enable
adoption easier. SMS/IVR advisory.
• Language + voice support ensures accessibility in Punjab • Data quality:-Use a hybrid model: fetch soil data from
and other regions. govt + allow farmer photo upload + guided form entry.
4
 IMPACT AND BENEFITS
Impact Benefits
❑ The proposed solution will create a ❑ The solution benefits multiple
measurable impact: stakeholder:
• Higher Crop Yields:- Personalized crop and • For farmer
fertilizer recommendations increase ➢ Accurate soil-based crop selection.
productivity by 20-30%. ➢ Reduced input cost, higher profit margins
• Cost Reduction:- Prevent excessive ➢ Reliable advisory in their native language.
fertilizer/pesticides use, saving money for • For Government and NGOs/Economics
farmers. ➢ Batter implementation of sustainable
• Sustainability:- Promotes balanced use of farming programs.
chemicals, reducing soil and water ➢ Reduced subsidy on fertilizers due to
pollution. efficient use.
• Empowerment:- Farmers get scientific, ➢ Improved rural livelihood leading to food
real-time insight instead of guesswork or security.
unreliable shopkeeper advice. • For Environment
• Digital Inclusion:- Multilingual and voice ➢ Less chemicals run off into soil and water.
support ensure accessibility for farmer ➢ Balanced soil fertility management.
with low literacy. ➢ Long-term soil health improvement.

5
 RESEARCH AND REFERENCES

❑ Market price tracking research and data ❑ Scientific Research Papers

• [Link] • Journals on precision agriculture, AI in farming,
various-commodities-various-markets-mandi and soil fertility management
• IEEE, Springer, Elsevier papers on AI-based crop
recommendation systems.

❑ For soil texture and land use survey

• The Soil and Land Use Survey of India (SLUSI) ❑ Government & Agricultural Institutes
➢ [Link] • ICAR (Indian Council of Agricultural Research)
:– Fertilizer and crop recommendations
• Punjab Agricultural University : – Soil health,
crop advisory guidesKrishi Vigyan Kendra
❑ Technology References
(KVK) publications
• TensorFlow / PyTorch for ML crop recommendation
modelsGoogle Cloud Vision API (for OCR in soil
reports)
• Firebase/[Link] for real-time app backend