Skin in the Game -- At the Collaborative Safety Research Center, Rini Sherony helped patent an artificial skin for test mannequins to accurately determine skin reflectivity. Photos by Brian Watkins
For nearly 20 years, Rini Sherony has lived by one truism: Work to create safer mobility for all road users and eventually make a lasting difference in the automotive industry.
Early on in her career, Sherony knew she wanted to be involved with safety research, which led to inventions that could change people’s lives in positive ways. When Toyota created its Collaborative Safety Research Center (CSRC) back in 2011, Sherony was asked to lead a collaborative team to create cutting-edge technological research.
“I love being involved in next-generation development research,” says the senior principal engineer. “When I moved to CSRC, my research became external and my findings were being published for the first time. I enjoy working in active safety and focusing on how to help people.”
Based in Ann Arbor, Michigan, the CSRC collaborates with universities that have technical expertise when it comes to safer mobility. Couple that with the practicality Sherony and the research team bring to the table, and the CSRC, delivers fresh, innovative ideas resulting in a number of patents.
Every Measurement Counts
One of the areas of research Sherony worked on, which led to a patent, is a method for enhancing lane departure assistance. This method warns drivers if their vehicle is about to leave a current lane and enter an adjacent one. The warning could be adjusted by measuring the driver’s pupil size and comparing it to baseline data stored in a driver profile to see if the driver is drowsy or not.
Let’s say a driver begins to doze off. The data stored within the car reads the driver’s pupil dilation size and can detect if the driver is getting drowsy. If it determines that the driver may be drowsy, an alert indicates that the chances of the car veering off into another lane are higher.
“I enhanced the method that not only uses lane marking, but detects drivers’ pupil sizes,” she says. “So, if the car goes across the lane, and the driver’s pupil doesn’t indicate drowsiness, it won’t give a warning right away. An alert can be issued at an appropriate situation based on the driver’s drowsiness status.”
Pedestrian crashes are increasing in the United States and in regions such as Asia and Europe. As a result, Sherony and her team set out to create a pedestrian detection method from video images, which can help notify a driver of an oncoming pedestrian before a collision occurs. This idea also has resulted in a patent.
The research for this image methodology required a significant amount of data collection and analysis on factors such as running and walking speeds, and distance of the person from the car. Once those parameters were extracted from the video, Sherony and her team could improve pedestrian detection research.
“My personal goal was to do naturalistic, real-time analysis of pedestrians because a large percentage of fatalities in Japan are pedestrians, and I noticed over the last five years that pedestrian fatalities were increasing in the U.S.,” she says. “We are facing more health challenges attributed to lack of movement, so more people are walking and biking to better their health. As such, more exposure is leading to more pedestrian crashes.”
The pedestrian detection issues a warning in real-time by using a camera-based system to monitor a driver’s surroundings. To do this, the system analyzes the video to generate a pedestrian detection score and signals an alert if the score is outside the predetermined threshold. This gives the driver time to react before a crash happens.
Safe and Sound -- Sherony’s personal interest in safe mobility works hand in hand with her next-generation development research at the Ann Arbor, Michigan, center.
Realistic Pedestrian Mannequins