Eunhan Ka

Postdoctoral Researcher · Purdue University

I develop physics-informed and data-driven methods to model, estimate, and secure traffic dynamics across large-scale transportation networks.

My research combines artificial intelligence, optimization, and network modeling to estimate transportation network states from incomplete observations, and to understand how the reliability of information shapes traveler behavior and network dynamics.

I am a postdoctoral researcher in the Lyles School of Civil and Construction Engineering at Purdue University, where I completed my Ph.D. under Dr. Satish V. Ukkusuri. My dissertation, Physics-Informed Neural Networks for Secure Connected and Autonomous Traffic Modeling, works at the intersection of traffic flow theory, machine learning, and behavioral modeling.

I study how transportation networks behave under disruption and adversarial manipulation (with emphasis on network resilience, transportation cybersecurity, and large-scale state estimation) and how reliable estimation and decision support can be maintained under uncertainty.

Education

Ph.D. Lyles School of Civil and Construction Engineering, Purdue University Dec 2025

Dissertation: Physics-Informed Neural Networks for Secure Connected and Autonomous Traffic Modeling
M.S. Department of Civil and Environmental Engineering, Seoul National University Feb 2018
B.S. Department of Civil and Environmental Engineering, Seoul National University Feb 2016

Research

Four focus areas, built on a shared foundation in network traffic dynamics.

hover or tap a focus area

News

Oct 2026 Received a Postdoctoral Supplemental Travel Award from the Purdue Office of the Vice Provost for Graduate Students and Postdoctoral Scholars, supporting travel to the 2026 INFORMS Annual Meeting. The award was made to 18 of 48 applicants across 11 Purdue colleges.
Sep 2026 Three papers are scheduled for presentation at the 106th Transportation Research Board Annual Meeting in Washington, D.C., covering origin-destination demand identification in retained-route Wardrop models, finite-grid used-path regret certificates for restricted traffic assignment, and context-matched scenario families for autonomous driving testing.
Sep 2026 Our paper on observation-aligned two-stage domain decomposition for physics-informed traffic state estimation is published in Transportation Research Part C: Emerging Technologies. The method reconstructs freeway speed fields from sparse fixed sensors while preserving shockwaves.
Sep 2026 Our paper on career-stage-specific training priorities for state department of transportation workforce development is published in Transportation Research Record, drawing on survey responses from 386 Indiana DOT employees.
Aug 2026 Presented MIRAGE, our work on detecting fake emergency electronic brake light attacks in V2X networks, at USENIX VehicleSec ‘26 (4th USENIX Symposium on Vehicle Security and Privacy) in Baltimore.
May 2026 New article in Transportmetrica B: Transport Dynamics on the resilience of traffic networks to route guidance attacks and the role of driver behavior heterogeneity.
Apr 2026 Gave an invited talk, Physics-Informed AI for Transportation Systems, in CGN 6425: Applied Data Science in Civil and Environmental Engineering, taught by Prof. Xiang Yan at the University of Florida.
Jan 2026 Started as a Postdoctoral Researcher in the Lyles School of Civil and Construction Engineering at Purdue University.
Dec 2025 Completed my Ph.D. in the Lyles School of Civil and Construction Engineering at Purdue University. Dissertation: Physics-Informed Neural Networks for Secure Connected and Autonomous Traffic Modeling.
Dec 2025 Our review, Cybersecurity for Next-Generation Road Transportation, is published in the ACM Journal on Autonomous Transportation Systems.
Nov 2025 New article in Transportation Research Part F on route guidance attacks in cyber transportation networks and behavioral sensitivity.
Jun 2025 Received the 2025 ITE Great Lakes District Student Paper Award.