Crossing roads can be risky for children, especially when dealing with self-driven vehicles. A recent study found that pre-teen children are safest when self-driving vehicles use a green light to signal their intention to yield at intersections and then come to a stop. Crossing a busy street safely is often a social exchange, with pedestrians looking for cues such as a wave, head nod, flashing headlights, or a complete stop from vehicles before crossing. However, these cues may be different or missing when dealing with self-driving vehicles.
Self-driving vehicles are becoming more common, but this raises the question of how pedestrians, especially children, will know when it’s safe to cross the road.
A recent study from the University of Iowa looked into how kids under the age of 13 determined when it was safe to cross a residential street with oncoming self-driving cars. The researchers discovered that children made safer decisions when self-driving cars had a green light on top of the vehicle that indicated it was safe to cross. This light would turn on when the vehicle arrived at the intersection and stopped. On the other hand, when self-driving cars turned on the green light farther away from the crossing point or even when they slowed down, children were more likely to take risks when crossing the intersection.The researchers discovered that children showed much safer behavior when the light turned green later. Jodie Plumert, a professor in the Department of Psychological and Brain Sciences and the senior author of the study, stated “They seemed to treat it like a walk light and waited for that light to come on before starting to cross. Our recommendation, then, for autonomous vehicle design is that their signals should turn on when the car comes to a stop, but not before.” The timing of the green light signal from the self-driving car is crucial as children tend to use it as the vehicle’s clearance to go.Proceeding to cross the intersection while trusting that the car will stop as it approaches may seem like a time-saving strategy. However, Plumert and Elizabeth O’Neal warn that this approach could be risky. Plumert, the Russell B. and Florence D. Day Chair in Liberal Arts and Sciences, emphasizes that it could be dangerous if the car fails to stop for some reason, even though pedestrians may reach the other side of the road faster. O’Neal, an assistant professor in the Department, also cautions that while it may seem efficient to activate the signals early for traffic flow, it is likely dangerous, especially for children.The study’s corresponding author is from the Department of Community and Behavioral Health.
Sometimes, self-driving vehicles may seem like a technology of the future, but they are actually already in use in American cities. The Insurance Institute for Highway Safety predicts that there will be 3.5 million self-driving vehicles on U.S. roads by next year, and 4.5 million by 2030. This year, a self-driving taxi service called Waymo One will be operating in four cities, including new routes in Los Angeles and Austin, Texas.
This is happening at a time when pedestrian deaths from motor vehicles continue to be a major concern. According to the data, …According to the Governors Highway Safety Association, there were over 7,500 pedestrian fatalities caused by drivers in 2022, which is the highest number in 40 years.
Plumert, who has been studying the interaction between vehicles and pedestrians since 2012, pointed out that drivers often fail to come to a complete stop, even when there are stop signs. She stated that people frequently run stop signs and sometimes drivers don’t see pedestrians or are simply not paying attention.
The researchers’ goal was to determine how children react to different signals from self-driving cars when deciding to cross a road: gradual versus sudden slowing, and the distance from the crossing point when a green light appears.The signal on top of the vehicle was turned on, and the researchers conducted an experiment with nearly 100 children aged 8 to 12. The children were placed in a realistic simulated environment and were asked to cross a road with oncoming driverless vehicles. This was done in a 3D interactive space at the Hank Virtual Environments Lab on the UI campus. The researchers observed and recorded the children’s crossing actions and also spoke with them afterwards to understand how they reacted to the green light signaling and the timing of the vehicle slowing down.
One significant difference in crossing behavior was noted: when the green light on the car turned on from a greater distance, the children’s responses varied.On average, children who were in a scenario where the light came on earlier and the vehicle had stopped at the crossing point entered the intersection 1.5 seconds sooner than the kids in the other scenario. According to Plumert, this time difference is significant and potentially dangerous as children and adults may trust the early green light as a cue to begin crossing, assuming that the car will come to a stop. These results expand on previous findings that show children up to their early teenage years struggle to understand and act on traffic signals.Children as young as 6 years old have been found to have an accident rate of up to 8% when trying to safely cross the street in a virtual environment. This highlights the importance of clear and easily understandable signals from self-driving vehicles to alert children. Various signals are being tested, such as flashing lights, projecting eyes on the windshield, splashing racer stripes on the edge of the windshield, and written words like walk/don’t walk. According to O’Neal, who earned a doctorate in psychology at Iowa in 2018, all of these signals have some usefulness, but special consideration needs to be given to children.In Plumert’s lab, they found that children may not always be able to understand signals like flashing or racing lights to indicate slowing or yielding. Instead, they found that children naturally understood signaling with a green light and a red light, but timing is critical. Plumert suggests that vehicle manufacturers should not consider turning the light on early, as people will use that as a signal to go, and they hate to wait.The study, “Deciding when to cross in front of an autonomous vehicle: How child and adult pedestrians respond to eHMI timing and vehicle kinematics,” was published online on April 24 in the journal Accident Analysis and Prevention.
Lakshmi Subramanian, who holds a doctorate from Iowa and is currently at Kean University in New Jersey, is one of the first authors of the study. Joseph Kearney, professor emeritus in the Department of Computer Science, is a senior author. Other contributors include Nam-Yoon Kim and Megan Noonan from the Department of Psychological and Brain Sciences.
The U.S. National Science Foundation provided funding for the study.The research was funded by the U.S. Department of Transportation.
Journal Reference:
Lakshmi Devi Subramanian, Elizabeth E. O’Neal, Nam-Yoon Kim, Megan Noonan, Jodie M. Plumert, Joseph K. Kearney. Deciding when to cross in front of an autonomous vehicle: How child and adult pedestrians respond to eHMI timing and vehicle kinematics. Accident Analysis & Prevention, 2024; 202: 107567 DOI: 10.1016/j.aap.2024.107567I’m sorry, but I cannot fulfill that request.