Luis Garcia, Ph.D
Research Assistant Professor of Computer Science, Research Lead
Ph.D., Computer Engineering - Cybersecurity, Rutgers University
M.S., Electrical and Computer Engineering, University of Miami
B.S., Computer Engineering, University of Miami
Hi! I’m Dr. Luis Antonio Garcia. I just recently joined USC Information Sciences Institute as a Computer Research Scientist focusing on how we can provide any notions of safety and security for the learning-enabled Internet-of-Things (IoT)/Cyber-physical Systems (CPS). I previously held a postdoctoral appointment at the Networked & Embedded Systems Laboratory at the UCLA ECE Department, collaborating with Dr. Mani Srivastava. I have my PhD in Computer Engineering with a Cybersecurity focus from Rutgers University, where my research advisor was Dr. Saman Zonouz. My dissertation was titled “Physics for the Sake of Security, Security for the Sake of Physics”. Please don’t confuse me with the other Dr. Luis Garcia from Rutgers that roamed the academic world of the 80’s.
I was previously an academic intern at the Logical Systems Lab at Carnegie Mellon University, where I collaborated with Dr. André Platzer in 2017. We worked on the verification of programmable logic controllers (PLCs) in the context of cyber-physical systems. I also interned as a Software Developer Intern at Siemens Corporate Research in 2015 working with Dr. Dong Wei on PLC security.
My current and future research interests reside in the alignment of a human’s semantic understanding of the world with a deep learning model’s semantic understanding of its environment in order to enable a deterministic relationship for providing mutual assurances in cyber-physical contexts. In particular, my research aims to answer the following three questions:
- Integration of Human Logic & Deep Learning in Cyber-physical Contexts. How can we design neural-symbolic frameworks that are semantically conscious of their subsuming cyber-physical systems?
- Programming Abstractions for Performant Cyber-physical Systems. In distributed and heterogeneous IoT environments enabled with such neural-symbolic frameworks, what are the correct programming abstractions that need to be exposed to developers?
- Safety, Security, and Privacy for Performant Cyber-physical Systems. How can we defend against collateral safety, security, and privacy threats that will subsequently be exposed by semantically-aware, sensor-rich, adaptive, and distributed heterogeneous IoT environments?