Abstract
Tracking the spatio-temporal movements of agricultural pests and diseases on a large scale is essential for informed and proactive management in precision agriculture. However, the traditional methods rely on manual data collection and analysis. These approaches are costly, time-intensive and impractical for long-term studies on a large scale. To address this gap, the present study introduces AgriTrack, a framework designed to assist in the tracking of pest and disease movement, perform temporal analysis, and predict farmers' information demand related to target pests or diseases. Leveraging fourteen years of extensive "Kisan Call Centre" data encompassing over 36 million farmers' queries from across India, the AgriTrack framework utilises a multidimensional approach. The proposed system offers functionalities including AI-based intra-seasonal and inter-seasonal spatial movement tracking, critical date shift extraction, and Deep Learning-based forecasting of farmers' information demand. The AgriTrack framework offers a scalable, data-driven approach that can reduce the time and resources associated with manual pest and disease monitoring. In addition, the study demonstrates the framework's real-scenario application through a case study focusing on tracking the spatio-temporal dynamics of the "Termite" pests in "Wheat" crop throughout India. The study showcases the framework's outputs, verifying its efficacy in providing actionable insights for agricultural stakeholders.