N.C. A&T’s Computational Data Science and Engineering Faculty Member Awarded Second Patent

EAST GREENSBORO, N.C. (Dec. 14, 2021) – Hyoshin “John” Park, Ph.D., an assistant professor in the Department of Computational and Data Science at North Carolina Agricultural and Technical State University, has been awarded his second patent, “System and Method for Predicting Effects of Forward Glance Durations on Latent Hazard Detection.” Recently, Park received his first patent for a cost-effective solution to reduce traffic congestion.

The patented system addresses two issues that have not previously been combined: detecting road hazards while simultaneously determining if the driver has lost focus on the road. This approach is particularly relevant to society’s safe transition to an autonomous future.

One possible commercial application is a safety system that uses a camera to monitor the driver’s eye motions along with a series of cameras or sensors that detect external hazards. A chime, beep, or flash of light will alert the driver both to refocus on the road and when it detects a hazard.

Park acknowledges that drivers frequently encounter distractions while on the road.

“Performance of in-vehicle, secondary tasks require a driver to alternate their glances between the inside of the vehicle and the forward roadway,” he said. “While previous research has shown that thresholds of off-road and durations are critical to latent hazard detection, there has been no research to predict the probability of hazard detection in a time series considering all possible forward glance durations.”

The patented system is the result of a study of 45 subjects testing eight computer-based scenarios in which they glanced at objects or other distractions both inside and outside of the vehicle during a driving simulation. When the tests concluded, Park used machine learning to measure whether the hazard detection likelihoods are correctly predicted per the different scenarios tested.

“A microbenchmark approach based on Hidden Markov models is explored to infer the transition probability of hazard detection that changes dynamically between the detection and non-detection stages. The model is cross-validated and demonstrated to be accurate and robust,” said Park.

This technology can be an important contribution to the automotive industry as it increasingly employs mixed-autonomous safety features.