Authors:
N. Anand, Edwin Shalom Soji

Addresses:
Department of Computer Science, Bharath Institute of Higher Education and Research, Chennai, Tamil Nadu, India.

Abstract:

This study proposes an innovative hybrid framework that combines deep learning architectures with EAI to enable adaptive crop recommendations. The system can collect data from various Internet of Things (IoT) sensors that provide real-time environmental data, such as soil moisture, temperature, humidity, and nutrient levels. The main goal is to provide a way to connect the complex model, implemented as a black box, with the transparency needed for decision-making in agriculture. For the research, a set of selected data comprising 286 instances across various soil and climatic conditions is used to train a hybrid CNN-RNN. The framework uses local interpretable model-agnostic explanations to explain the rationale behind individual crop recommendations. The tools include libraries of deep learning algorithms (Python), specialised kits for sensor integration, and data visualisation platforms. The results show that the hybrid approach is not only effective in terms of predictive accuracy but also builds farmers' trust by providing information on the effects of different environmental factors. This multimodal sensing and explainable modeling are a step forward in precision agriculture, enabling the sustainable use of resources and maximizing crop yields across different climatic conditions.

Keywords: Precision Agriculture; Deep Learning; Explainable AI; Multimodal Sensors; IoT Sensors; CNN and RNN; Visualisation Platforms; Crop Recommendations; Climatic Conditions.

Received on: 14/04/2025, Revised on: 19/06/2025, Accepted on: 04/09/2025, Published on: 05/06/2026

DOI: 10.69888/FTSESS.2026.000716

FMDB Transactions on Sustainable Environmental Sciences, 2026 Vol. 3 No. 2, Pages: 82-91

• Views : 57
• Downloads : 7

Download PDF