Applied AI

AI on real hardware: computer vision and imitation learning, so a robotic arm can learn simple tasks from demonstrations.

What we do

Applied AI connects machine learning to physical engineering, and to the Robo Arm platform in particular. Rather than treating AI as software alone, the project applies computer vision, data collection, sensor feedback, imitation learning and task planning to real hardware.

Members work on camera-based perception, object detection, dataset collection, teleoperation, robot demonstrations, model training, evaluation and live task execution. The long-term goal is a system where a robotic arm learns simple tasks from demonstrations and performs them reliably.

Who's on it

  • Fahad Mehmmood

    Project Lead · Applied AI

  • Muhammad Tayyeb Bukhari

    Project Lead · Applied AI

  • Usman Waseem

    Project Lead · Applied AI

  • Ahmed Yar Khan

    Member · Applied AI

  • Hayan Azhar

    Member · Applied AI

  • Meraal Rahman

    Member · Applied AI

  • Muhammad Aayan Abdullah Malik

    Member · Applied AI

  • Zarnab Tariq

    Member · Applied AI