Constructed an autonomous self-driving car using computer vision, leveraging lane detection, YOLO-based object recognition, and low-latency processing pipelines for real-time obstacle avoidance and decision-making.

Invented a marine sensing system integrating an edge-AI module into autonomous profiling floats. The system enables real-time underwater acoustic event detection using TinyML and onboard semantic translation of complex NetCDF data into accessible summaries for cloud platforms.

Invented an intelligent LPG management system integrating IoT sensors, TinyML edge intelligence, and ensemble machine learning for predictive refill alerts and automated offline leak detection. Secures supply chains through blockchain smart contracts and builds user trust with Explainable AI (XAI).

Invented an intelligent, IoT-driven vermiculture automation system utilizing ESP32 microcontrollers and a hybrid sensor array to optimize bio-fertilizer production. Introduces a novel Worm Health Mapping Index (WHMI) for real-time biological stress calculation alongside automated environmental actuation and cloud-based mobile connectivity.

Invented a cost-effective, autonomous AI-powered shopping cart system tailored for high-density retail environments. Features smartphone-based BLE/WiFi localization for movement, checklist-driven aisle navigation via path-planning algorithms, and integrated RFID/barcode scanning for instant self-checkout using offline-capable edge computing.

Invented a robust, GPS-free traffic speed monitoring system utilizing Ultra-Wideband (UWB) sensor nodes. Employs a self-healing, mesh-based Two-Way Ranging (TWR) protocol for sub-microsecond synchronization, achieving precise, weather-immune vehicle speed detection in complex urban environments without relying on LIDAR or radar.

Developed a GAN-based deep learning framework for Parkinson's Disease detection, generating synthetic medical images to address data scarcity and improve classification performance.

A research-driven weather forecasting system developed to analyze large-scale meteorological data using PySpark's distributed computing framework. Integrated datasets from IMD, AWS, and ARGS to generate accurate weather predictions and analytical insights.

Proposed a novel disk scheduling algorithm that dynamically calculates the optimal scanning direction using spatial proximity and request density. Demonstrated through simulations to outperform SSTF in average wait time and resolve starvation variance for heavy I/O workloads.

Invented a closed-loop healthcare supply chain ecosystem integrating causal AI for predictive demand forecasting and autonomous procurement. Features edge AI-enabled robotics for secure, network-resilient last-mile delivery of medical supplies, backed by a blockchain audit layer for an immutable chain of custody.

Invented an adaptive, multi-entropy authentication framework utilizing dynamic expander graph traversal for OTP generation. Replaces deterministic structures with session-specific topological entropy amplification, integrating biometric fusion and adaptive risk-scoring to mitigate replay vulnerabilities and computational threats.