As an undergraduate studying computer science and engineering in New Delhi, Rohan Paul, the son of two doctors, took on a class project working with the blind. He learned that the white cane, which many visually impaired people depend on for mobility, has a major shortcoming: It doesn’t allow users to detect obstacles at or above knee level, such as tree branches, road barriers, car doors and clotheslines.

“Each person had a story of how they were hit on the head,” he says. Parents of blind children “were scared of sending their children to school because they were afraid they’d smack their heads.”

So back in 2005, Dr. Paul and his classmates developed a prototype device to help blind people detect such obstacles. In the years since, he says, “this extracurricular class project became a mission.”

After a Rhodes scholarship at Oxford, where he earned a Ph.D. in mobile robotics, Dr. Paul returned to India and worked on refining the device. It is a small, handle-shaped attachment for the top of a cane, with an ultrasonic sensor that perceives taller obstacles that the cane alone would miss. The polycarbonate, battery-powered attachment vibrates when a user approaches an obstacle—the closer the object, the faster the vibrations. At a distance of up to 8 feet, it can detect an object as small as just over an inch across.

When potential users told Dr. Paul, 31, and his colleagues what they ideally wanted, they cited ease of use, safety and affordability. They also sought to avoid stigma. “People wanted something so subtle,” he says. Men wanted to avoid bumping into women on the street—and the resulting humiliation. Women wanted something small and light. The goal, Dr. Paul says, was “safe, independent mobility with dignity.”

When he tested the device from 2012 to 2014 with nearly 200 people in six cities, users reported 90 percent fewer upper-body collisions than with an ordinary cane.