PhD Researcher · Civil & Environmental Engineering (Transportation)

Mohammad
Elayan

I develop intelligent, human-aware autonomous driving for mixed traffic, centering on multi-objective driving, behavioral trade-off balance, and driver heterogeneity modeling. My work draws on empirical trajectory analysis, safety analysis, microsimulation, geospatial analysis, optimization, and AI/ML methods, including reinforcement learning and quantum-inspired techniques.

Curriculum Vitae ↗ Research ↓ Get in touch ↗
Mohammad Elayan

Transportation & Autonomy Laboratory for Intelligence
University of Nebraska-Lincoln

§01 · About

Mohammad Elayan is a PhD Researcher in Civil Engineering (Transportation) at the University of Nebraska-Lincoln and a member of the Transportation and Autonomy Laboratory for Intelligence, directed by Dr. Wissam Kontar.

His research develops intelligent, human-aware autonomous driving for mixed traffic, centering on multi-objective driving, behavioral trade-off balance, and driver heterogeneity modeling. His work draws on empirical trajectory analysis, safety analysis, microsimulation, geospatial analysis, optimization, and artificial intelligence and machine learning methods, including reinforcement learning and quantum-inspired techniques. His research has featured in top journals and conferences.

Before his doctorate, he spent nearly a decade as a Senior Transportation Engineer in Dubai, leading numerous transportation projects across the MENA region, including transportation master plans, traffic impact studies, and multimodal integration for major developments and city-scale planning efforts.

Multi-objective driving Driver heterogeneity Trajectory analysis Simulation & optimization AI & machine learning

§02 · Research

Research Interests

Four threads run through the work below, each paired with a figure drawn from, or in the style of, the underlying research.

Fig. 1 · Consensus

Behavioral Consensus for Automated Vehicles

Quantifying how automated vehicles balance safety, interaction quality, and efficiency in mixed traffic, using empirical Pareto frontiers and reinforcement learning to steer behavior toward better trade-offs.

Empirical Pareto frontier for behavioral consensus

Fig. 2 · Heterogeneity

Quantum-Inspired Driver Heterogeneity

Representing each driver as an evolving, probabilistic behavioral state, a density matrix learned without supervision, so that heterogeneity and its interactions emerge directly from naturalistic trajectory data.

Quantum-inspired driver heterogeneity representation

Fig. 3 · Trajectories

Empirical Trajectory & Safety Analysis

Mining large-scale naturalistic and simulated trajectories for interaction and conflict measures, combining street-view, LBS, and crash data to model risk with finer spatial and behavioral resolution.

Empirical trajectory and safety analysis

Fig. 4 · Optimization

Microsimulation, Optimization & RL

Calibrating microsimulation environments to real-world traffic and using reinforcement learning to steer AV/HDV behavior toward system-level, consensus-aware objectives.

Microsimulation, optimization, and reinforcement learning

§03 · News

Latest News

Aug 27, 2026

New Preprint: Quantum-Inspired Modeling of Driving Behavior

Our paper "Quantum-Inspired Modeling of Driving Behavior," submitted to Transportation Research Part B, represents each driver as an evolving density matrix that is continuous, probabilistic, and history-dependent, with individual behaviors aggregating to recover known traffic phenomena such as the fundamental diagram and hysteresis.

Paper ↗ Code ↗

May 25, 2026

Published in Transportation Research Part C

Our paper "Can Automated Vehicles Have It All? A Consensus Framework for Diagnosing Behavioral Trade-offs in Mixed Traffic" introduces a framework for quantifying behavioral consensus, evaluating how well traffic behavior satisfies safety, interaction, and efficiency objectives across six metrics.

Paper ↗ Code ↗

May 3, 2026

Accepted at IEEE ITSC 2026

Our paper "Behavioral Heterogeneity as Quantum-Inspired Representation" was accepted for IEEE ITSC 2026 in Naples, Italy (Sept 15 to 18), introducing a quantum-inspired framework for representing behavioral heterogeneity in intelligent transportation systems.

Preprint ↗ Code ↗

Feb 9, 2026

Lightning Talk at NAIRR Annual Meeting 2026

Our talk "Consensus-Aware AV Behavior: Empirical Pareto-Guided Reinforcement Learning for Mixed Traffic" was accepted to the NAIRR Pilot Annual Meeting 2026, supported by NSF computational resources for training consensus-aware RL models at scale.

Jan 15, 2026

Accepted at IEEE IV 2026

Our paper "Learning the Pareto Space of Multi-Objective Autonomous Driving: A Modular, Data-Driven Approach" was accepted for IEEE IV 2026 in Detroit, a timestep-based, data-driven framework for consensus-aware driving with no access to future data.

Preprint ↗ Code ↗

§04 · Connect

Get in Touch

Contact

melayan2@nebraska.edu

WHIT 362E
Prem Paul Research Center at Whittier School
Lincoln, Nebraska, 68503

Academic Profiles

Google Scholar ↗ ORCID ↗ ResearchGate ↗ LinkedIn ↗ GitHub ↗

Research Interests

  • Behavioral Consensus for Automated Vehicles
  • Quantum-Inspired Driver Heterogeneity
  • Empirical Trajectory & Safety Analysis
  • Microsimulation, Optimization & Reinforcement Learning