Recent research:
The researcher combines affordable sensing, computer vision, and machine learning to create accessible, autonomous systems with tangible societal impact. Across projects, they develop low-cost, locally processing AI solutions for assistive technologies (auditory traffic-light status for the visually impaired using a Raspberry Pi with transfer-learned CNNs), mobile robotics for safe mine-detection and real-time ground mapping on uneven terrain, immersive VR/AR educational tools for children with Down Syndrome using VR and Leap Motion with locally trained neural networks, and AR-enabled predictive maintenance for industrial equipment (Weibull-based reliability analysis, wireless sensor networks, and cloud-enabled data fusion via MQTT). Collectively, the work emphasizes accessible hardware, real-time perception and decision-making, human-centered design, and the integration of AI/ML with IoT and cloud services to enhance safety, learning, and maintenance in resource-constrained settings. The methodology spans transfer learning, local edge computation, robotics navigation and mapping, VR/AR interface design, neural network training with local data, and statistical reliability modeling, with impact evident in improved detection accuracy, safety, education, and maintenance efficiency.