As a result of funding by the Toyota Economic Loss Settlement, researchers are trying to determine if a crash can be predicted well in advance, partly based on the driver’s increased stress levels.
Details of the current research, led by the Texas A&M Transportation Institute (TTI), were released during the closing session of the 2015 Traffic Safety Conference.
“The purpose of the Toyota project is to identify different methods and technologies that we can use inside vehicles to detect vehicle and driver based errors,” said Mike Manser, manager of TTI’s Human Factors Program. “We want to be able to predict an increased risk of these two scenarios well in advance so that we can provide either control to the vehicle or information to the driver to reduce the propensity for or to avoid a crash.”
Working on the project are researchers from TTI, Texas A&M University’s Departments of Industrial and Mechanical Engineering, and Engineering Technology and Industrial Distribution; the University of Houston Computational Physiology Laboratory; Texas Tech University; and the University of Michigan Transportation Research Institute.
So far, 48 participants have driven a stress-inducing course on TTI’s driving environment simulator. Using heart rate detectors, thermal cameras, and sensors able to detect slight increases in perspiration, researchers evaluate stress levels compared to driver performance.
“The data we have looked at so far are already showing good evidence that the type of technology we thought was science fiction really is useful for detecting driver stress,” Manser told conference attendees.
In other phases of the project, participants will drive actual vehicles on a closed course at TTI’s Riverside Campus as researchers examine driver stress and vehicle-based errors and then provide on-board notifications or vehicle control as a safety countermeasure.
The $3 million, multi-year research project was made possible as a result of the 2013 Toyota Economic Loss Settlement.
For more information, contact Mike Manser at [email protected].