Research Projects

Year 1: Foundation and Data Integration - The initial year will focus on building a comprehensive multi-source data ecosystem for Odisha. Remotely sensed imagery, ground-truth agricultural data, meteorological records, and disaster archives will be integrated to create a unified repository. The gap addressed here is the fragmentation of datasets that currently limits robust modeling. Early prototypes of uncertainty-aware AI models using granular computing and fuzzy logic will be developed to process ill-defined/ indiscernible and overlapping data regions. This establishes the baseline for reliable agricultural and disaster analytics. Year 2: Model Development and Uncertainty Management - The second year will advance into rough-fuzzy hybridization and deep learning models to capture complex spatio-temporal dynamics. The focus will be on uncertainty management, improving the interpretability and trustworthiness of AI outputs. Novelty lies in combining granular deep learning with rough-fuzzy methods, an area scarcely explored in the Indian context. Prototype systems will be tested for disaster risk mapping and preliminary agricultural land classification (1-crop vs 2-crop). Year 3: Agentic AI Framework and Agricultural Land Management -The third year will deliver the agentic AI framework, capable of autonomously integrating heterogeneous data streams while maintaining human oversight. The system will support interpretable decision-making for agricultural land management, mapping soil health, irrigation potential, and sustainable conversion of 1-crop to 2-crop systems. This addresses the critical gap of lack of interpretable AI tools for farmers and policymakers. Conservation strategies for fertile 2-crop lands will also be introduced, ensuring food security. Year 4: Energy–Agriculture–Disaster Integration - In the fourth year, the research will expand to multi-energy hubs (electricity–heat–water) powered by AI to optimize rural energy for irrigation, cold storage, and disaster preparedness. This represents a novel convergence of AI-driven energy optimization with agricultural resilience, uplifting the research capacity of the department and projecting the institute (IIIT Bhubeneswar) as a leader in sustainable technology solutions. Year 5: Validation, Policy Integration, and Scalability - The final year will emphasize large-scale validation, stakeholder engagement, and policy integration. Demonstrations in Odisha’s coastal and drought-prone districts will prove scalability. By ensuring explainable and trustworthy AI, the framework will align with governance needs and serve as a replicable model for other Indian states. The outcome of the proposed research will strengthen the research stature of IIIT Bhubaneswar, positioning it at the forefront of AI for disaster resilience and sustainable agriculture.