IJMRR - International Journal of Multidisciplinary Research & Reviews
Int. J. Multidiscip. Res. Rev.
DOI Prefix: 10.56815
I J M R R
Int. J. Multidiscip. Res. Rev.

International Journal of Multidisciplinary Research & Reviews

Peer Reviewed | Multidisciplinary | UGC Aligned | C | O | P | E Ethical Standard

ISSN: 2945-3135 | Digital Archiving | CrossRef Member
IJMRR - International Journal of Multidisciplinary Research & Reviews

International Journal of Multidisciplinary Research & Reviews

Peer Reviewed · Multidisciplinary · UGC Aligned · C | O | P | E Ethical Standard

ISSN: 2945-3135 | Digital Archiving | CrossRef Member
DOI DOI Prefix: 10.56815
Vol. 2025 No. 2: 2025 (2025)

[24] Precision Farming with AI and ML: A Comparative an Alysis of Mysuru District Villages and Surroundin G Areas Using Npk and Ph Sensor Data

Noor Fathima , Dr. Vinay Kumar S B & Dr. Mohmad Umair Bagali

With the aid of artificial intelligence and sensor technology, precision farming presents encouraging answers to the problems facing contemporary agriculture. This study uses soil fertility data from portable NPK and pH sensor modules to compare precision farming methods in Mysuru district villages with those in neighboring Karnataka, India. Crop yield data for rice, maize, sugarcane, and ragi were recorded together with environmental factors like temperature, humidity, soil moisture, and light intensity. Long Short-Term Memory networks were used to anticipate rainfall, Random Forest was used to predict yield, and Support Vector Machines were used to classify soil fertility. The average rice yield in Mysuru was 1200 kilograms per acre, according to the results. This was around 20% more than the paddy yields in Chamarajanagar, where productivity was constrained by soil acidity and nitrogen shortage. In Hassan, nutrient imbalance decreased yields to 1100 kilograms per acre, whereas in Hunsur, under nutrient-rich conditions, maize produced 1500 kilograms per acre. Due in significant part to improved potassium availability and sensor-guided fertilizer application, Mysuru's sugarcane fields recorded over 38 tons per acre, which is around 6–8 tons more than Mandya's average. Thanks to more consistent levels of soil moisture and nitrogen, ragi yields in Mysuru were 1000 kg/acre as opposed to 800 kg in Hassan. With soil fertility classification at 92% and yield prediction R2 = 0.89 in Mysuru, machine learning models outperformed those in neighboring locations, which had R2 = 0.74 and 78%, respectively. According to surveys, 35% of Mysuru farmers have embraced soil sensors, IoT irrigation, or mobile apps, whereas less than 15% of neighboring villages have done the same. These results highlight the need to extend precision farming methods to underserved rural areas while also showing that combining AI and ML with inexpensive soil sensor data can greatly increase productivity and sustainability. 

Keywords:
Precision Agriculture Soil Sensors NPK Monitoring Machine Learning Random Forest SVM LSTM Crop Yield Prediction Karnataka India