Kisaan Vaani

Inspiration

Indian farmers face frequent crop losses due to unpredictable weather, soil variability, and late detection of crop diseases. Traditional soil testing labs are costly, time-consuming, and impractical for regular use, while most digital solutions lack local language support and actionable guidance. Kisaan Vaani was inspired by the need to give farmers a simple, affordable, and reliable way to make data-driven decisions directly from their fields.

What it does

Kisaan Vaani is an IoT- and AI-powered smart agriculture platform that provides real-time soil analysis, crop recommendations, and crop disease detection. Using custom-built sensors, it measures soil parameters such as moisture, pH, temperature, and nutrients. Machine learning models analyze this data to suggest the top five crops best suited to the farmer’s soil and investment capacity. Farmers can also upload crop images to detect diseases and receive organic and inorganic treatment recommendations through a multilingual mobile app.

How we built it

We designed a compact IoT device using an ESP32 and custom-built soil sensors, with NPK sensing integrated externally. Sensor data is transmitted to Firebase in real time. Machine learning models were trained on self-curated datasets containing 95,000+ soil records and images of 150+ plant diseases. A Flutter-based mobile application was developed with support for 13 Indian languages, integrating cloud services for data storage, model inference, and user notifications.

Challenges we ran into

Building low-cost yet reliable sensors suitable for real field conditions was a major challenge. Collecting and cleaning high-quality datasets for both crop prediction and disease detection required significant effort. Ensuring accurate predictions under varying lighting, soil, and weather conditions was also complex. Additionally, designing an intuitive multilingual interface that farmers could easily adopt required multiple iterations.

Accomplishments that we're proud of

We successfully built a fully functional end-to-end system—from hardware to AI models to a user-friendly mobile app. The project uses custom datasets and sensors, reducing dependency on expensive labs. The disease prediction module achieves high accuracy on real-world images, and the crop recommendation system delivers personalized, location-specific suggestions. Supporting 13 Indian languages makes the solution inclusive and accessible.

What we learned

We gained hands-on experience in integrating IoT hardware with cloud platforms and AI models. The project taught us the importance of dataset quality, field testing, and user-centric design. We also learned how technology adoption improves when solutions are affordable, simple, and tailored to real user needs.

What's next for Kisaan Vaani

Future plans include integrating weather forecasting, yield prediction, and pest outbreak alerts. We aim to enhance sensor accuracy, add satellite data integration, and expand disease coverage. Partnerships with government bodies, FPOs, and agri-input companies can help scale deployment and bring Kisaan Vaani to more farmers across India.

Built With

• api
• cpp
• iot
• machine-learning
• microcontroller
• python