Nineteen-year-old Romanian Ionut Budisteanu has won top prize at the Intel International Science with a driverless car and a small host of webcams.
Budisteanu, who recently graduated from high school, said that he’d wanted to create a driverless car ever since 2004, when he learned that 2.1 million people, worldwide, died in car accidents that were caused by driver error.
While companies like Google and Audi have been pumping millions into driverless car programs, Budisteanu couldn’t exactly afford that kind of price tag. After looking at how Google was engineering their driverless cars, though, he realized that you could make a cheaper system by replacing the expensive, high-resolution 3D radar that was the cornerstone of the program.
To lower the cost of the system, Budisteanu employed a lower-resolution 3D radar, one which was perfectly sufficient for dealing with large objects such as “other cars” and “buildings.” While this was mostly sufficient, there was still a need to capture the finer details – and that’s where the webcams come in. To process this dizzying amount of data, Ionut employed artificial intelligence algorithms to extract information from the visual data and plot a safe route for the car.
Even running on this low-cost, multi-unit, home-made 3D radar system, Budisteanu was able to achieve excellent results. In fact, the vehicle performed flawlessly in 47 of 50 simulations, only failing to recognize some people at a distance of 20 to 30 meters.
For Budisteanu’s efforts, he was awarded $75,000, and a Romanian company has already offered to fund his research and give him the opportunity to create a prototype driverless car.
Watch a Video Budisteanu’s Car in Offline & Online (15:00) Simulations:
Image & Video Courtesy of Intel & Ionut Budisteanu