Cavon H. ’27 Improves Pedestrian Safety With Invention of Hindsight
In the United States, pedestrians are injured in a traffic crash roughly every seven minutes. Cavon H. ’27 is no stranger to this experience. Walking on the street, he was nearly hit by a car at age 13. He searched for a device that could warn pedestrians about cars approaching from behind, but discovered no such product existed. Rather than give up, he decided to build one himself.
The experience led to the creation of Hindsight, a wearable radar device designed to give pedestrians early warning of approaching vehicles. This spring, the project earned international recognition at the Regeneron International Science and Engineering Fair (ISEF) in Phoenix, Arizona, the world's largest pre-college science competition.
Worn around the waist, Hindsight uses a custom 10 GHz Doppler radar system to detect vehicles approaching from behind and alert the user through a mobile app and vibration feedback. The device can detect vehicles up to 100 meters away and provides an average of seven seconds of warning time. The project evolved significantly from the prototype Cavon brought to science fairs a year earlier.
“Last year’s prototype was very much a proof of concept,” he said. “Although it worked in parking lots, it didn’t really work well in real-world environments, and so I wanted to expand and make it something that’s actually feasible, but still very low cost.”
To reach that goal, Cavon designed his own radar hardware, signal-processing board, algorithms, and machine-learning model. He estimates that he went through approximately 18 hardware iterations before arriving at the current design.
One of the biggest challenges was reducing false alarms. Radar signals are affected by the wearer's own motion, surrounding pedestrians, and multiple vehicles traveling at different speeds. Cavon developed a layered system combining traditional signal processing with machine learning to distinguish genuine threats from background activity.
“You don’t want it constantly going off,” he said. “So I had to go through a lot of custom algorithm design and machine learning to really bring down those false positives.”
Accessibility was also central to the design. The device runs for more than 20 hours on a single charge, and Cavon estimates the manufacturing cost could eventually be reduced to less than $10 per unit.
“My end result is not necessarily to create a good science fair project,” he said. “It’s to create a sensor that I can give to people, and they can really use to keep themselves safe.”
Hindsight's performance earned Cavon first place in Electronics and Electromagnetics at the Los Angeles County Science and Engineering Fair before advancing to the California Science and Engineering Fair, where he again placed first in his category.
At ISEF, which featured more than 1,700 finalists from 67 countries, Cavon received three awards: Second Place Grand Award in Embedded Systems ($2,400), First Place Innovator Award sponsored by Qorvo ($5,000), and First Place Association of Old Crows Award for electromagnetic innovation ($2,500). He was the highest-placing U.S. finalist in the Embedded Systems category.
Beyond the awards, Cavon said one of the most meaningful aspects of ISEF was the opportunity to meet fellow young scientists. “You get to meet so many people from all over the world who are so different from you,” he said. “You all kind of have that shared interest in engineering and science and making things that matter.”
Although much of the technical work took place at home, Cavon credits Poly faculty and resources with helping launch the project. Upper School Science Teacher Robin Barnes introduced him to science fairs, while Upper School Science Teacher Jack Prater, whose background includes antenna and high-frequency design at JPL, provided guidance on radar design and engineering challenges.
“Mr. Prater gave me a lot of guidance and advice in terms of what directions I could take the project and technical advice on radar design, and Ms. Barnes introduced me to the whole science fair system,” said Cavon. “I wouldn’t have had these opportunities had it not been for Poly’s Upper School Science Department.”
Looking ahead, Cavon hopes to continue testing Hindsight with runners and pedestrians and to use award funding to further refine the device.
“My end goal is to create something people will actually use,” he said. “That's really the target.”