Selected Publications
A curated archive of peer-reviewed articles and manuscripts. Rather than a purely academic list, this catalog highlights the industrial translation and applied value of each research contribution.
Autonomous Driving, VLM, and AI Safety
Context-Aware Conformal Prediction for VLM-Based Driving Scene QA
YJ Joo, Y Kim, SY Park, M Abdel-Aty
Applies conformal prediction to calibrate vision-language models (VLM) for autonomous driving scene question answering, guaranteeing reliable confidence bounds in safety-critical settings.
Provides safety-critical AI systems with calibrated uncertainty estimation, reducing hallucination risks in autonomous vehicle decision-making.
MMCAformer: Macro-Micro Cross-Attention Transformer for Traffic Speed Prediction with Microscopic Connected Vehicle Driving Behavior
L Han, M Abdel-Aty, Y Kim, YJ Joo, Z Islam
Introduces a transformer architecture (MMCAformer) integrating macroscopic traffic patterns and microscopic driving behavior from connected vehicle telemetry.
Provides traffic managers and AV navigation systems with highly accurate, short-term speed prediction capabilities under dynamic traffic conditions.
Generative artificial intelligence for class imbalance in crash severity estimation with mixed data types
EJ Kim, DK Kim, HS Lee, YJ Joo
Leverages generative AI (like tabular GANs) to generate synthetic minority-class crash samples, resolving data imbalance issues in crash severity prediction models.
Enhances the predictive accuracy of insurance risk models and safety analytics tools where severe crashes are historically underrepresented.
Predictive Safety-Aware Transmit Rate Control Scheme for Real-Time Proactive Forward Collision Warning System
YJ Joo, DK Kim, EJ Kim
Proposes a V2X transmit rate control scheme that dynamically adjusts the communication rate based on predictive safety risks to prevent forward collisions.
Improves wireless resource utilization in connected vehicles while ensuring zero safety compromises for real-time safety critical systems.
Are Autonomous Vehicles Perceived as Riskier than Human Drivers?
YJ Joo, J HWANG, M Abdel-Aty
Explores public perception of risk regarding autonomous vehicles vs. human drivers, utilizing empirical survey data to model and quantify perceived risk.
Helps AV companies and policymakers design trust-building campaigns and address public anxiety regarding AV safety.
Transportation Safety and Smart Mobility
Micro-level hotspot identification at intersections using traffic conflict analysis
N Park, J Park, YJ Joo, M Abdel-Aty
Utilizes video-based traffic conflict metrics to detect micro-level intersection safety hotspots before crashes actually occur.
Enables cities to conduct proactive infrastructure safety audits, saving millions in emergency safety repairs.
Propagation of Positional Measurement Errors in Traffic Conflict Analysis
YJ Joo, Z Islam, M Abdel-Aty, DK Kim
Formulates how GPS/video tracking positional errors propagate into calculated traffic conflict indicators like Time-to-Collision (TTC).
Provides error-correction bounds for computer vision teams developing automated safety monitoring systems.
Queue Length-Based Priority Strategy for Unsignalized Intersections in Mixed-Traffic Environments
BYH Kim, YJ Joo, J Choi, DK Kim
Develops a priority strategy for autonomous and connected vehicles at unsignalized intersections, optimizing queue lengths and minimizing delays.
Supports the design of traffic coordination systems for mixed autonomous/human driver environments.
Effects of heterogeneity in driving manoeuvres on calibrating car-following models at a roundabout
J Choi, YJ Joo, DK Kim
Investigates how diverse driving styles and roundabout configurations affect the calibration and reliability of car-following traffic models.
Improves the fidelity of micro-simulation tools used by traffic engineers to evaluate new road designs.
A generalized driving risk assessment on high-speed highways using field theory
YJ Joo, EJ Kim, DK Kim, PY Park
Applies artificial potential field theory to model dynamic driving risk on high-speed expressways, capturing interactions between the vehicle, road geometry, and traffic flow.
Serves as a foundation for real-time collision-avoidance systems and adaptive cruise control safety wrappers on highways.
A data-driven Bayesian network for probabilistic crash risk assessment of individual driver with traffic violation and crash records
YJ Joo, SY Kho, DK Kim, HC Park
Develops a probabilistic Bayesian network to assess individual driver crash risk by combining historic traffic violation records and crash demographics.
Enables auto insurance underwriters and commercial fleet managers to build personalized risk profiles and identify high-risk drivers.
Reliability-based assessment of potential risk for lane-changing maneuvers
YJ Joo, HC Park, SY Kho, DK Kim
Introduces a reliability theory approach to quantify the instantaneous crash risk during lane-changing maneuvers, accounting for speed and spacing uncertainties.
Provides safety-critical thresholds for ADAS lane-keep assist systems to prevent side-swipe and rear-end collisions.
Injury severity of bus–pedestrian crashes in South Korea considering the effects of regional and company factors
HC Park, YJ Joo, SY Kho, DK Kim, BJ Park
Explores commercial transit safety by analyzing the multi-level effects of transit companies and regional environments on bus-pedestrian crash severity.
Helps municipal transit authorities identify systemic safety flaws and optimize driver training programs.
Intersection Crash Frequency Analysis Considering Visual Environment Features Using Random Parameter Negative Binomial-Lindley Model
L Han, M Abdel-Aty, YJ Joo, S Zhai, D Wang
Analyzes how visual clutter, signage, and environmental factors influence the frequency of intersection crashes using random parameter statistical models.
Informs civil design on optimal signage placements to reduce driver cognitive load and crash rates.
Large-Scale Traffic Data and Signal Analytics
Arterial Network Traffic State Prediction with Connected Vehicle Data: An Abnormality-Aware Spatiotemporal Network
L Han, M Abdel-Aty, YJ Joo
Uses connected vehicle telemetry to build a spatiotemporal deep learning model capable of predicting arterial network congestion under non-recurrent anomalies.
Provides advanced traffic routing engines with real-time congestion forecasting during accidents or special events.
Evaluating the Effect of Signal Adjustment on Rear-End Conflicts Using High-Resolution Event-Based Data
YJ Joo, M Abdel-Aty, BMTH Anik, Z Islam
Leverages high-resolution event-data (ATSPM) to model how traffic signal timing adjustments impact rear-end conflict rates at signalized intersections.
Allows city traffic managers to tune signal timings to simultaneously optimize flow and safety.
Order dispatching in ride-pooling with walking points search
SY Park, YJ Joo, DK Kim
Proposes a ride-pooling dispatching algorithm that dynamically routes passengers to nearby walking pickup/dropoff points to maximize system efficiency.
Helps ride-sharing platforms (like Uber/Lyft) reduce vehicle miles traveled (VMT) and operational costs while improving passenger wait times.
Estimating Network-Wide Arterial Traffic Volume and Speed under Sparse Sensing Using Connected Vehicle Data
S Yang, M Abdel-Aty, YJ Joo, L Han
Models network-wide arterial traffic volumes and speeds by fusing sparse fixed detector data and low-penetration connected vehicle traces.
Lowers data-collection costs for departments of transportation by utilizing commercial CV data instead of physical loop detectors.