At Meghna Group of Industries, I investigate electrical, automation, and instrumentation failures through root-cause and performance analysis. I also support commissioning, sensor calibration, and automation improvement at Meghna Pulp & Paper Industry Ltd., MGI.
I proposed FAHRNet, a frequency-aware hybrid deep-learning architecture for ×4 remote-sensing image super-resolution, designed to recover high-frequency spatial details while preserving structural image quality. My research also includes uncertainty-aware inference for Earth observation and area-based control for voltage safety in distribution systems.
For an academic project, I developed a low-cost embedded impact-testing system for PVC pipes, integrating strain-gauge sensing, RP2040-based data acquisition, and real-time analysis. My work draws on electrical engineering and computer science, including machine learning, image processing, and industrial control.
