Research 

Current Projects (under construction) 

Context-aware Coordinated Multi-agent Reasoning for Safe Robot Operations, part of " AI CRAFT: Artificial Intelligence Cybersecurity Readiness and Future Training", US DoD, DARPA. 24-26


Robots are now ubiquitously employed across various settings, ranging from autonomous vehicles and warehouse logistics to challenging rescue missions, performing myriad tasks with minimal human guidance to no human intervention at all. Fundamental to these robotic operations is scene perception - the capability to perceive and understand the surrounding environment, mainly reliant on vision sensors such as cameras and LiDAR systems. This aspect is paramount for decision-making, navigation, and, eventually, successful task execution.

However, in complex environments, such as uncontrolled road intersections or multi-robot coordination scenarios, distributed agent reasoning, such as semantic segmentation, can sometimes falter, lacking situational awareness or leading to adverse outcomes during adversarial attacks on one of the agents. We aim to investigate the influence of centralized and ad hoc coordination in multi-agent robotic perception under the dual challenges of maintaining situational awareness and dealing with adversarial attacks. 

Completed projects -  Under Construction

iDriveSense, was part of an NSERC Strategic grant

iDriveSense integrates vehicular and smartphone sensors within and amongst vehicles to offer robust road monitoring, driver profiles and route recommendations. The collected data is then filtered, processed, and sent to a road information system (RIS) cloud where it is used to build road and driver information repositories. iDriveSense transforms the vehicle into a smart node capable of providing vital information for road safety. We remark that our solutions are not intended to replace existing RIS installations and navigation systems, but rather provide such systems with robust real-time information that can significantly improve the services they offer including route-planning and navigation services. The different components of iDriveSense and the new design of route selection criterion based on driver preferences, supported by road conditions monitored and reported in real-time are explored.

A. Abdelrahman, A. S. El-Wakeel, A. Noureldin and H. S. Hassanein, "Crowdsensing-Based Personalized Dynamic Route Planning for Smart Vehicles," in IEEE Network, vol. 34, no. 3, pp. 216-223, May/June 2020.

A. S. El-Wakeel, J. Li, A. Noureldin, H. S. Hassanein, and N. Zorba "Towards a Practical Crowdsensing System for Road Surface Conditions Monitoring", IEEE Internet of Things Journal, vol. 5, no. 6, pp. 4672-4685, Dec. 2018. 


Selected Publications 

Journals:

Conferences:

* Authors contributed equally.