Research

Current Research Activities

  1. Human-Machine Coadaptation Based on Reinforcement Learning with Policy Gradients.
    Researcher: Karim Tahboub

    The problem of adaptive human-machine interaction is investigated. It is sought that not only the human learns how to perform a task with a novel machine, but the machine itself co-adapts to the human style in the interaction. This requires solving the problem of two agents co-adapting or co-learning at the same time. Due to the lack of human learning and performance models, it is hypothesized that reinforcement learning with policy gradient algorithms are good candidates for addressing this problem with robustness and fast convergence

  1. Islamic ethics as basis for ethical AI.
    Researcher: Karim Tahboub

    In designing artificial moral agents, ethical challenges are encountered such as AI biases, fairness, human dignity, responsibility and liability, and privacy. For this, ethical theories stemming from Western moral philosophy such as those classified under meta ethics, applied ethics, and normative ethics are dominantly referred to. In this research, Islamic ethics as presented in the holy Quran is explored as an alternative framework for AI ethics.

  1. Multi-body formulation as an alternative modelling approach to mobile robots subject to motion constraints.
    Researcher: Karim Tahboub
    The mainstream of kinematic and dynamic modelling of mobile robots subject to motion constraints is based mainly on differential geometry concepts including vector fields and accessibility distributions. Multibody mechanics formulation and in particular the concept of modes of motion offers an alternative approach to develop the kinematic and dynamic models, to find a reduced model, and to represent the system in state space. Further issues as presence of friction, contact with moving object, and control can be addressed efficiently using this formulation.

 
Publications:

  • Maharmeh, E. and Tahboub, K. A. (2023). Human Machine Co-adaptation using Co-Adaptive Policy Gradient Algorithm. 7th International Conference on Control, Automation, and Diagnosis (ICCAD)
  • Iqnaibi, E., AbuShokor, A., Alsaied Ahmad, L., and Tahboub, K. A (2023). A Robotic Soil Excavator for Truck Loading: Kinematics, Dynamics and Motion control for a 4-DOF Robotic Manipulator Prototype. 7th International Conference on Control, Automation, and Diagnosis (ICCAD)
  • Tahboub, K. A. (2022). Modes of Motion as an Alternative Approach for Mobile Robot Kinematic Constraint Representation. 8th International Conference on Control, Decision, and Information Technologies (CoDIT), pp. 962-966
  • Tahboub, K. A. (2019). Human-Machine Coadaptation Based on Reinforcement Learning with Policy Gradients. 2019 8th International Conference on Systems and Control (ICSC), 2019, pp. 247-251, doi: 10.1109/ICSC47195.2019.8950660.
  • Bader, A., Sharawi, A., Saeed, M. and Tahboub, K. A (2017).  A Novel Three Degree of Freedom Stewart-like Platform for Testing Humanoid Postural Capabilities. Australian Journal of Basic and Applied Sciences 11(115), pp. 113-125.
  • Abdalrahim, I., Ben Abdessmad, J., Anwar, M., Cuneo, M., Hendrik, H., and Tahboub, K. A. (2017). South-South Cooperation for Science Diplomacy, ITEC Programme on Science Diplomacy, Research and Information System for Developing Countries, India, pp. 59-64

 
 
Master Theses:

  • Maharmeh, E. (Jan. 2023). Reinforcement learning-based human-machine co-adaptation via policy gradients.
  • Amer I. (Feb. 2023). Using Deep Pose Estimation of Football Player Body for Virtual Reality.
  • Al-Qais, M. (Jan 2023). Single Board Computer ROS-Based Tennis Balls Collecting Mobile Robot.
  • Attawna, M. (Jan 2020) Using Reinforcement Learning to Learn Seega Board Game.
  • Al-Mohtaseb, M. F. (ongoing). Tennis Ball Collecting Robot.