November 16, 2023 | Updated: November 15, 2023
Around the world, there are more than 1.3 million fatal car collisions each year. The total includes vulnerable road users such as cyclists, pedestrians and motorcyclists. IEEE Member Bilin Aksun-Guvenc believes that these numbers can be significantly reduced, and pedestrian safety increased, with a device that most people carry in their pockets: the smartphone.
Aksun-Guvenc’s research uses smartphone data to inform drivers and autonomous vehicles of collision risks.
Tell us about the system you are working on.
Smartphones are equipped with an array of sensors — including GPS, accelerometers, gyros, step counters and light sensors — that can measure a vulnerable road user’s motion. With a smartphone app, this data can pinpoint a user’s location, future position and travel direction, then broadcast it to vehicles via an over-the-air connection.
Why wouldn’t carmakers just use machine vision to identify pedestrians and other vulnerable road users?
Carmakers do use machine vision with existing vehicle cameras for protection, but these systems can’t see around corners or through barriers since they need a line-of-sight to detect users. Vision sensors are also affected by lighting and weather conditions; for instance, some lane-keeping assist systems deactivate in the rain. Communication systems, less hindered by these conditions, could complement machine vision to improve user tracking.
So, does the pedestrian need a special app on their smartphone? Does the driver need anything? Could this become a standard feature in cars?
The same app could operate on the driver’s smartphone as well as on the vulnerable road user’s device. Since many users already have smartphones, integrating this technology into production vehicles would be straightforward. This system, especially with level 2 automation vehicles, could significantly enhance collision avoidance measures.
How does the user’s phone communicate with the vehicle?
Low-energy Bluetooth is a viable option due to its minimal power consumption. It allows for ”advertising mode,” wherein the phone broadcasts data without pairing, maintaining user privacy. That said, the technology is agnostic with regard to communication technology. Cellular technologies could also serve this purpose without significant latency impacts.
Would it be possible to incorporate this system into devices that aren’t phones?
Absolutely, and it’s recommended. Construction workers could carry beacons similar to digital safety vests, while cyclists could incorporate the system into GPS devices. This simple beacon approach would benefit children who don’t own smartphones.
What challenges have you faced in designing this system?
Real-world deployment with many users is necessary to refine our approach and evaluate its full potential. Issues like smartphone GPS accuracy and communication latency can lead to false positives and negatives, but these are resolvable problems.
Are there any other applications for this system?
The system’s communication features could aid in various scenarios, such as guiding food delivery robots on university campuses which often block access ramps while waiting to cross streets. Better vehicle communication could mitigate this. Additionally, vehicle-to-vehicle cooperation could enhance urban mobility, potentially replacing physical traffic signals at intersections.
Learn more: What makes an autonomous vehicle safe? How safe should we expect them to be? The IEEE Standards Association has worked to establish best safety practices for autonomous vehicles based on car following, road sharing, pedestrian situations, intersection scenarios and visibility issues including consideration of rules of the road and their regional and/or temporal dependencies.