Current Projects
Innovation Engine Project: Data-Driven Mapping of U.S. Innovation Deserts
Research Fellow – Southern Methodist University, Texas (NSF-Funded)
This project focuses on identifying and benchmarking “innovation deserts” across 3,222 U.S. counties to inform regional policy interventions. Leveraging Kernel PCA (KPCA) with Gradient Boosting Machines (GBM), the work has achieved an 11% reduction in model training cost and a 15% improvement in anomaly detection AUC through synthetic data generation and hybrid model development.
Key outputs include a carbon-penalizing objective function for renewable energy integration optimization and dashboard-driven KPIs to guide prioritization across 65 U.S. metro areas.
Network Efficiency and Security Optimization in Next-Generation Wireless Systems
Research Fellow – Edge Hill and Lancaster Universities, UK
This project enhances heterogeneous IoT system performance by combining hardware–software optimization techniques and designing AI-driven security solutions for smart grid and communication networks.
Research also involves the development of AI and digital forensics curricula to integrate emerging technologies into electrical engineering education.
Completed Projects
Electricity Theft Detection and Attack-Resilient Power Grids
Multiple funded projects (NSF USA, EPSRC UK, British Council, EU SANCUS)
Worked on a series of funded projects to develop stacked machine and deep learning models for electricity theft detection, achieving 23% improvement in predictive accuracy.
Developed robust frameworks for attack-resilient power grids, including heuristic algorithm-based OPF models incorporating stochastic renewable energy sources.
Key Publications:
- IEEE Transactions on Power Systems, 2022
- IEEE Transactions on Smart Grid, 2022
- IEEE Transactions on Instrumentation and Measurement, 2025
- Elsevier Energy Reports, 2025
Big Data Analytics for Smart Grid Optimization
Funded by EPSRC & British Council
Developed big data analytics frameworks for:
- Short-term load forecasting in smart grids.
- Stochastic renewable energy integration.
- Optimal power flow solutions using Augmented Grey Wolf Optimization.
These models improved forecasting accuracy and reduced computational costs, enabling scalable deployment in large-scale power networks.
Control and Forecasting in Renewable Energy Systems
Collaborations with industry and academic partners
Designed adaptive nonlinear MPPT control for mitigating intermittency in offshore wind power systems.
Worked on short-term electric load forecasting using novel stacked ensemble frameworks such as Kolmogorov-Arnold Networks (KANs).
Nuclear Robotics and Autonomous Inspection
Funded by EPSRC National Centre for Nuclear Robotics
Contributed to developing autonomous inspection algorithms for nuclear fuel manufacture, integrating AI-based control and optimization into robotic systems for responsive and sustainable nuclear operations.