Skip to main content
Suat Gumussoy

Suat Gumussoy

Associate Professor - Electrical & Computer Engr
 
+1 (972) 883-2623
ECN 3.518
Closed-Loop AI Lab
Google Scholar
LinkedIn
ORCID

Currently accepting students

Professional Preparation

Ph.D. - Electrical and Computer Engineering
The Ohio State University - 2004
M.S. - Electrical and Computer Engineering
The Ohio State University - 2001
B.S. - Mathematics (Double Major)
Middle East Technical University - 1999
B.S. - Electrical and Electronics Engineering
Middle East Technical University - 1999

Research Areas

Closed-loop AI for cyber-physical systems:
  • Physics-informed digital twin modeling
  • AI-driven control (multi-objective reinforcement learning, explainable AI)
  • AI-driven design (Bayesian optimization, generative methods)
with applications to power systems, manufacturing, and edge healthcare. 
Broader interests include system identification, robust (H-infinity) and time-delay control theory, and reinforcement learning.

Publications

Unsupervised Pretraining for Neural Value Approximation - Journal Article
Timings for Numerical Experiments on Benchmark Examples for Fixed Order H∞ Controller Design - Journal Article
Computing ℋ∞ norms of time-delay systems - Conference Paper
Computing H-infinity Norm of Time-Delay Systems - Journal Article
A Bottom-Up Approach for Searching for Sparse Controllers with a Budget - Journal Article
Inverter-based resource (ibr) optimized fault-level adjustment based on fault location 2025 - Other
A bottom-up approach for searching for sparse controllers with a budget 2025 - Journal Article
Conceptual closed-loop design of automotive cooling systems leveraging Reinforcement Learning 2025 - Journal Article

Appointments

Senior Key Expert on Closed-Loop AI
Siemens Research [2019–2026]
Senior research scientist driving closed-loop AI research directions including fundamental contributions, software framework/toolbox implementations and their applications to industrial engineering problems.
Principal Developer in Controls and Identification Team
MathWorks [2011–2019]
Lead architect and developer of Reinforcement Learning Toolbox. Key contributions to Robust Control, Control System and System Identification Toolboxes.
Postdoctoral Associate
KU Leuven [2008–2011]
Postdoc on fixed order robust controller design for time delay systems using non-smooth, non-convex optimization methods. Developed a MATLAB-based toolbox implementing several numerical methods and controller design algorithms.