News

The paper proposes an ML-assisted framework that combines fast IR sensor-based track segment detection with PID gain scheduling for competitive line-following robots. The approach achieves a 100% completion rate on composite test tracks using a Neural Network scheduler, outperforming fixed-gain baselines. A participation and presentation certificate was awarded by the ICPRAM 2026 Secretariat.

This paper presents a novel approach to accelerating track segment detection in competitive line-following robots using IR sensor array data and machine learning. The proposed method combines physics-based simulation with ML-assisted PID gain scheduling to achieve real-time adaptive control without replacing the classical PID backbone. The work draws on experience mentoring ACM student competition teams.

The student team, working under my mentorship through the Robotics and AI Club at the American College of the Middle East, built and tuned a competitive line-following robot. The team integrated concepts from C programming, digital design, and microcontrollers courses into their competition robot, demonstrating strong sensor integration, PID control tuning, and robust autonomous behavior on the competition track.

Responsibilities include curriculum oversight, faculty coordination, and enhancing student academic experiences within the CET program. This role builds on my experience teaching across the full range of CET courses since August 2024.

The ceremony marked the formal recognition of my doctoral work on equalization in optical fiber communication using model-based neural networks, conducted within the EU H2020 MSCA REAL-NET network at Télécom Paris. A proud moment shared with family.