Ohio State teams with Ford to advance pedestrian safety

Posted: December 8, 2022
Smartphone app displays a warning on driver's windshield
The smartphone app displays a warning which is reflected on the driver's windshield.

Researchers from The Ohio State University are helping make roads safer for pedestrians, cyclists, scooter riders and drivers alike with a new smartphone app developed as part of a research alliance with Ford.

The technology, developed by Ohio State’s Automated Driving Lab (ADL), helps warn motorists of approaching pedestrians, cyclists and more, even those blocked from a driver’s view. The app running on a pedestrian’s phone uses Bluetooth Low Energy (BLE) messaging to communicate their location to nearby drivers who are also running the app. If the app calculates a potential crash risk, the driver is alerted with audio and visual warnings.

Data from the National Highway Transportation Safety Administration estimates traffic fatalities increased 13 percent to 7,342 in 2021 versus the prior year. Bicyclist traffic fatalities increased 5 percent to 1,000 during the same time period.

Ohio State’s ADL is led by Professors Levent Guvenc and Bilin Aksun-Guvenc. Since most road users often have a mobile phone with them, the ADL team’s Android app uses the multitude of sensors available in these devices to collect motion data and communicate with other mobile phones or a cloud server using internet connectivity.

“The vulnerable road user data that we collect is used to train recurrent neural networks for future motion and intent prediction,” said Guvenc, a professor in mechanical and aerospace engineering, and electrical and computer engineering. “That is then broadcasted using Bluetooth Low Energy advertising or cloud communication to nearby vehicles where the driver’s mobile phone listens to this information and uses its open location and heading information to compute different levels of warning.”

Levent Guvenc (headshot)
Levent Guvenc

ADL researchers implemented the necessary filtering and recurrent neural network in the mobile phone app along with the communication capability. The mobile phone in the vehicle is placed in a heads-up display so that the driver can easily notice and hear the warnings. A permanent Bluetooth device is also used in the vehicle in some of the experiments to improve the range of communication.

The technology is easy to implement and is not vehicle-dependent, making it possible to see benefits immediately, according to Aksun-Guvenc, a research professor in mechanical and aerospace engineering. Ford is researching the technology further in the hope of one day providing warnings to the drivers in the presence of pedestrians, bicyclists and scooterists even in hard to see situations.

“All that is needed is the pedestrian safety app running in mobile phones of the vulnerable road users and also in vehicles,” she said. The app runs in the background, does not use or record any personally identifiable information, and does not drain unnecessary power from the battery.

The safety app also works in cases where there is non-line-of-sight, such as when the driver or the perception sensors of the car cannot see when a pedestrian is behind a building or parked vehicles. The app can easily be bundled with other apps used on university campuses or in cities and can be particularly useful for alerting drivers close to school zones during busy times of day. Additionally, the technology can be embedded in food delivery robots used on many university campuses, which are also at risk for collisions with motorists.

Bilin Aksun-Guvenc (headshot)
Bilin Aksun-Guvenc

Along with Guvenc and Aksun-Guvenc, the research team includes Sukru Yaren Gelbal, post-doctoral researcher in the ADL and a recent Ohio State PhD graduate, who performed the programming, algorithm development and experimental evaluation.

Future potential applications for the technology include vehicle-to-vehicle communication and cooperative driving, especially for fully electric vehicles used in urban environments. The technology can also be extended to advanced driver assistance systems to perform automated braking and steering to avoid possible accidents. For road users without access to a mobile phone, lower cost wearable devices or clothing with simple location and communication capability can be developed to achieve the same safety objective. These wearable devices can also be embedded into wheelchairs or canes for mobility or visually impaired vulnerable road users.

ADL researchers have been conducting applied research since 2010 on connectivity between all road users for improved safety, efficiency of traffic flow, and fuel and energy usage optimization. The in-house developed, connected and autonomous driving simulator at ADL has been a focus of inspiration for researchers within the university and beyond.

“Our applied connected vehicle and autonomous vehicle expertise, and past experience of many successful industry project collaborations have placed Ohio State in a unique position for this vulnerable road user safety project,” said Guvenc. “We are very excited about this technology and look forward to working with interested communities to improve the safety of vulnerable road users.”

by Meggie Biss, College of Engineering Communications | biss.11@osu.edu

Categories: ResearchFaculty