Embedded Files
Latest Update: Sep. 2026
Coordinating independent decision makers
When independent agents share airspace and infrastructure, a sensible choice for one can create congestion or risk for others. I study how information, incentives, and limited oversight can align individual decisions with safety, fairness, and system performance while preserving decentralized decision authority.
My work develops game-theoretic methods and studies their use in air traffic management and autonomous systems.
CORE THEME
Noncooperative coordination
Independent operators may share routes, airspace, or capacity while pursuing different objectives. Each decision changes what is feasible or costly for others. My research studies how recommendations, incentives, and limited oversight can make individually rational decisions support safer and more effective collective outcomes.
The methods are developed for specific coordination settings, with assumptions and guarantees stated for each model.
01 / INFORMATION DESIGN
Correlated recommendations
Correlated recommendations can coordinate independent agents by informing each about the likely actions of others. My work studies how to compute these recommendations at scale: reduced-rank correlated equilibria use a convex hull of selected Nash equilibria, while chance-constrained formulations address uncertain costs and provide probabilistic incentive-compatibility guarantees under stated assumptions.
Ongoing work investigates smaller joint-action representations for exclusive resource-assignment games.
02 / INCENTIVE DESIGN
Trading and negotiation
When agents prefer different outcomes, agreement may require changing the consequences of their choices. The Trading Auction for Consensus (TACo) lets agents negotiate over candidate outcomes using transferable assets, with finite-time convergence under its model assumptions.
Iterative negotiation and oversight adds a limited regulatory signal when negotiation alone leaves undesirable system-level shortfalls. A decentralized CTOP case study examines efficiency and fairness while leaving the final choice with stakeholders.
03 / REGULATORY PROTOCOL DESIGN
Rules and dynamics
Rules and oversight matter when information and incentives alone do not ensure required operational constraints. In airspace sector overload mitigation, I study regulated best-response updates and conditions for convergence and overload resolution. Earlier Nash-based free-flow air traffic work examined decentralized conflict resolution with limited communication.
Each model states the conditions under which its safety or convergence result applies.
THEORY & ANALYSIS
Trade-offs in noncooperative coordination
Decentralized decisions create tensions between what each agent will accept and what the system needs. Faster convergence can come at the expense of system performance; stronger oversight can limit autonomy; and protecting private preferences can restrict the information available for coordination. Fairness is itself a coordination objective: in some settings, I optimize it directly alongside aggregate cost and study how system gains are distributed across agents. I analyze these trade-offs within each model’s assumptions, including when safety, incentive compatibility, and convergence guarantees can coexist with efficient and fair outcomes.
APPLICATION AREAS
Air traffic management and advanced air mobility
Air traffic management and advanced air mobility provide concrete settings for strategic coordination. My work spans free-flow conflict resolution, sector overload mitigation, decentralized Collaborative Trajectory Options Program (CTOP) operations, and vertiport and landing-sequence coordination. In each setting, independent decision makers share airspace or capacity under safety constraints, with goals that include reducing delay and improving fairness.
CURRENT RESEARCH DIRECTIONS
Dynamic and resilient coordination
One direction asks how bounded incentive signals can steer a system between desirable operating equilibria while agents retain decision authority. I use dynamical-systems and optimal-control tools to study stability, feasible transitions, and intervention budgets. In parallel, I am studying chance-constrained correlated recommendations for noncooperative spacecraft collision avoidance and robust AAM landing-sequence coordination when agents may misreport preferences or act maliciously. Across these projects, the goal is reliable coordination despite strategic behavior and imperfect information.
EARLIER WORK
Multi-drone inspection routing
Earlier work developed scalable route-planning methods for cooperative multi-drone infrastructure inspection. The methods combined collision-free spatiotemporal planning with graph clustering for dense inspection sites, including problem instances with roughly 1,000–2,000 nodes. The video below illustrates this line of work.
Austin, Texas, United States
dlawogks2005@gmail.com
jaehan.im@utexas.edu
Page updated
Google Sites
Report abuse