AI-Powered Climate-Aware Crop Recommendation System using IoT and Machine Learning

1,2,3,4,5UG Students, Department of Electronics & Communication Engineering, Haldia Institute of Technology, Haldia, Purba Medinipur, West Bengal 6Faculty, Department of Electronics & Communication Engineering, Haldia Institute of Technology, Haldia, Purba Medinipur, West Bengal ABSTRACT Increasing climatic unpredictability has made traditional, experience-based crop selection increasingly unreliable for Indian farmers. This paper proposes an AI-Powered Climate-Aware Crop Recommendation System that combines IoT-based environmental sensing with Machine Learning to deliver real-time, data-driven crop recommendations. The proposed architecture spans an ESP32 microcontroller with DHT22 and soil sensors, a Node.js/PostgreSQL backend, a Random Forest Classifier trained on agricultural crop-recommendation data, and a Next.js dashboard that presents recommendations to the farmer. For this symposium demonstration phase, given hardware and time constraints, the crop-recommendation inference was simulated using Google's Gemini API with manually entered field parameter values in place of live sensor feeds, in order to validate the end-to-end recommendation logic and dashboard pipeline ahead of full hardware and model deployment. Sample simulation results demonstrate the system correctly identifying high-confidence crop matches from a given set of atmospheric and soil-nutrient parameters, indicating strong potential for real-world precision-agriculture deployment.