Speaker Bio
Murtadha Alsayegh is a Ph.D. candidate at the Knight Foundation School of Computing and Information Sciences (KFSCIS) at Florida International University (FIU), working on Security in Robotics. His current research interests are Secure Multiparty Computation (SMPC), Privacy-Preservation, and Secure Motion Planning, performed at the Motion, Robotics, and Automation Lab (MORA Lab). Before pursuing his Ph.D. program, Murtadha earned a Master’s degree in Software Engineering in 2012 at the University of Michigan-Dearborn and a B.S. in Computer Science at Lawrence Technological University in 2010. As a Ph.D. student, he was awarded a scholarship from the Saudi Arabian Cultural Mission (SACM). Murtadha also worked as a Senior Automation Engineer at Schneider Electric – Saudi Arabia, in 2013. He hardened several supervisory control and data acquisition systems against security threats and cyber-attacks in several major power plants. He has published two research papers at IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2020) and European Control Conferences (ECC2022).
Description
The robotics research community has made substantial progress in planning, controlling, and coordinating multi-robot systems in recent years. Robots have become core components of our everyday existence and routines, including home activities, manufacturing, healthcare, and surveillance system. These robots use a large amount of information for their proper functionality. However, often there are great potentials in which some robots will be curious to learn more information (e.g., semi-honest adversary) than necessary. Furthermore, in some scenarios (e.g., compromising sensitive information), robots can lead to significant disasters when they divulge sensitive information for malicious purposes. In these cases, we have to rely on methods that allow the robot to collaboratively work together to compute their parts privately without sharing unnecessary information while achieving the optimal purpose. This dissertation proposes a four-pronged approach to solving the privacy problem between heterogeneous robots.
First, we present Lightweight communication protocols for data synchronization without exchanging the initially required $n$ bits. Second, we propose a secure multiparty auction-based assignment algorithm in which we allocate robots to tasks securely without divulging any information (e.g., valuations, utilities, positions, or related data) between parties without the need to involve a third party (e.g., trusted auctioneer). Third, we propose a solution by employing the secure multiparty computation with Markov Decision Process (MDP) to solve a planning problem between parties in which they need to operate a robot to accomplish tasks. Our thrust four will involve a study case for heterogeneous robots that investigate multiparty computation to serve multiple teams in achieving communication cost-efficient, task allocation, and planning approaches while guaranteeing privacy to all parties.
Click here to join the Zoom meeting.
Meeting ID: 993 357 2674
Passcode: PD5436