Palo Alto Teen's Asteroid Detection System Wins STEM Award

PALO ALTO, CA — A Palo Alto teen’s innovation received grand recognition last month at the Regeneron International Science and Engineering Fair, winning the $10,000 Peggy Scripps Award for Science Communication and another $5,000 first-place award in the physics and astronomy category.

Franklin Wang, a 17-year-old junior at Palo Alto High School, developed a neural network that detects small asteroids near Earth about twice as well as current methods. Wang used machine learning and synthetic data sets to uncover six objects that had never been seen before.

Only 0.1 percent of small asteroids — which are less than 144 feet in diameter — are detected because they move fast and are too small to be seen. But Wang was able to train his software to identify streaks left on images by the objects, which were all less than 70 feet in diameter.

“The approach I came up with here is that my dataset is completely structured artificially,” Wang said in an interview with Patch. “So you can simulate the asteroid moving across that telescope exposure and smearing it out, and you can create a giant data set with all these asteroids.”

Wang examined four nights of images from the Palomar Observatory in Southern California and said he would like to continue his research with the Zwicky Transient Facility — an astronomical survey at the observatory — to apply his algorithm to more data.

In his freetime, Wang plays the oboe and the English horn. He also runs the student non-profit Project Code Foundation and has taught middle schoolers machine learning.

Wang plans to spend the money on his tuition for college, where he wants to do undergraduate research and take an interdisciplinary approach in computer science and astronomy.