These Teenage Girls Are Some Of The Most Promising Scientists Of The Future

Girls made up more than half of the finalists at Google’s Science Fair, the largest in the world. From turning banana peels into plastic to diagnosing skin cancer, here were a few of our favorite projects.

The Google Science Fair, now in its third year, is the biggest science fair in the world–simply by virtue of the fact that most of the competition takes place online. So it makes sense that the fair, open to kids between 13 and 18, also has some of the most diverse contestants of any science fair. The grand prize winner, Eric Chen, won for an impressive project that models new anti-flu drugs. But many of the best projects put forth by the 15 finalists came from girls, who made up over half of all the finalists (The nearly equal gender ratio among finalists was not intentional, says Mary Lou Jepsen, a judge in the competition and head of the display division at Google X Lab.)


I took a trip to Google HQ earlier this week to hear from the finalists. These are some of our favorites.

Producing Bioplastic From Banana Peels

Elif Bilgin, the 16-year-old Turkish winner of both the Voter’s Choice Award and the Scientific American Science in Action prize, won for her method of producing plastic from banana peels. Bilgin says that she was inspired to create biodegradable, fossil fuel-free plastic from seeing all the plastic trash floating in the Bosphorous strait, located in her hometown of Istanbul. When she discovered that other researchers were working on plastic made from potato starch, Bilgin decided to go a different route. “I wanted to use a waste material instead of something we consume,” she says.

So she went with banana peels, which are 30% starch. “In Turkey, we eat a lot of fruit,” she says. Bilgin’s bioplastic production recipe: dipping a banana peel in a special solution to prevent decay, boiling and pureeing it, adding a chemical to break down the starch, pouring the starch into a mold, and baking it. Universities in Turkey have already offered to help Bilgin continue her research. In the future, she hopes to study engineering.

Diagnosing Melanoma, No Human Help Necessary

Elizabeth Zhao, a finalist from Portland, Oregon, came up with a new way to diagnose melanoma using computers to analyze a processed image (a company called Mela Sciences has developed something similar that is now being used in doctor’s offices). Zhao’s inspiration came two years ago, when a friend’s mother passed away from melanoma. When she discovered how easy it is to cure melanoma that’s found early–you just slice it off–she set out to find a better way to diagnose the disease than the traditional method of visually surveying the skin.

Zhao’s process: capture an image of the mole, digitally remove the hair and extract key features, and then analyze the image for warning signs of melanoma, such as asymmetry, irregular border, abnormal color, and large diameter. The young scientist’s machine learning algorithm gets better at identifying melanoma when more data points are entered; with 350 data points, it’s about 80% accurate.

Zhao’s next step is to collect more images of melanoma to improve her system’s accuracy. In the past, she says, universities have been reluctant to give her what she needs because she’s only a high school student. “I hope the Google Science Fair gives me credibility,” she says.


A Better Way To Predict The Spread Of Tumors

Esha Maiti, a 17-year-old from the Bay Area, experienced a loss from cancer firsthand when her grandmother died from breast cancer. By the time doctors realized that her tumor had metastasized to her vital organs, it was too late. Why, Maiti wondered, did the doctors miss the tiny tumors spread throughout her grandmother’s body?

She soon realized that small tumors–those that are 10 to 15 mm wide–blend into tissue and are impossible to see using traditional imaging techniques. “If we knew the growth rates and metastasis rates of tumors, we could predict metastasis,” she explains. Growth rates vary depending on the person, type of cancer, and type of tumor, so Maiti developed an algorithm that can predict the size and number of metastasized tumors in the body based on the size of the primary tumor.

Maiti’s research is on hold, she says, while she finishes college applications. She hopes to study chemical engineering, computer modeling, or computer science.

Check out all the Google Science Fair winners and finalists here.


About the author

Ariel Schwartz is a Senior Editor at Co.Exist. She has contributed to SF Weekly, Popular Science, Inhabitat, Greenbiz, NBC Bay Area, GOOD Magazine and more