2018 National Winners

1st Place Grand Prize (Received iPad and engraved award) – Sneha Revanur (Quimby Oak Middle School - California)
2nd Place Grand Prize (Received Apple Watch and engraved award) – Gitanjali Rao (STEM School Highlands Ranch - Colorado)
3rd Place Grand Prize (Each student received a Samsung Galaxy tablet and engraved award) – Kailey Aponte, Janeliz Guzman, Bria Roettger (Ramey Unit School = Puerto Rico)
4th Place (Each student received an Apple TV and engraved award) – (TIE) Krisha Patel (Margaret Mead Junior High - Illinois); Mary Cotter (Allendale Columbia School - New York)
5th Place (Each student received a Fitbit and engraved award) – (TIE) Lara Scherer, Haruka Doi (Environmental Charter School at Frick Park, Pittsburgh Colfax K-8 - Pennsylvania); Nidhi Mathihalli, Elizabeth Stoiber (Redwood Middle School - California); Nivedita Sanghavi (Chaboya Middle School - California)

Honorable Mentions (Each honorable mention received a Google Home Mini and an engraved award): 

  • LeiLoni DePew, Taylor Dryden - (Pioneer Ridge Middle School, Wheatridge Middle School - Kansas)
  • Hannah Tandang - (Oxford Middle School - Kansas)
  • Sara Kaufman - (American Heritage School - Florida)
  • Megan Shetty, Nikita Jadhav - (Aviara Oaks Middle School, Double Peak K-8 School - California)
  • Riya Datla, Srinidhi Mallela - (Thomas Hart Middle School - California)
  • Sruthi Kurada - (Advanced Math and Science Academy Charter School - Massachusetts)
  • Riya Gupta, Stuti Agarwal - (The Harker School, Redwood Middle School - California)
  • Anushya Shankar - (Thomas Grover Middle School - New Jersey)
  • Samhita Gone, Sahithi Pogula - (Hopkinton Middle School, John F. Kennedy Middle School - Massachusetts)
  • Tara Adusumilli, Niki Surapaneni - (RJ Grey Junior High - Massachusetts)
  • Tania Mishra, Stephanie Hoffman, and Mallory McGuire - (Wisconsin Hills Middle School - Wisconsin)
  • Alicia Roice, Ashlyn Roice - (Altamont Elementary School - California)
  • Riya Jain - (Margaret Mead Junior High - Illinois)
  • Deepti Aggarwal - (Smith Middle School - Michigan)
  • Isabel Tucker, Amanda Gilpin - (Jefferson Middle School - Tennessee)
  • Sruthi Sankararaman (Rachel Carson Middle School - Virginia)

All finalists received a trophy and certificate at the 2018 ProjectCSGIRLS National Gala.


2018 Semifinalists

2018 Finalists


Finalists' Project Descriptions and Videos

Megan Shetty, Nikita Jadhav

Aviara Oaks Middle School, Double Peak K-8 School, California

pic1.jpg

 8 million metric tons of plastic enter the ocean each year, harming marine life and polluting the oceans. And even contributing to climate change. Biodegradable plastic had started to be used in order to help the plastic issue. However, it did have its own set of problems...for example, biodegradable plastic has acid in it, and if it goes to the ocean it would harm coral (acid is harmful to the coral) Biodegradable plastic also sinks, which limits its exposure to UV rays, not allowing it to break down in water. Therefore, after 1-2 months of research, we made plastic that broke down in water, and would be safe for marine life. Our original recipe included vinegar, however due to the fact that vinegar has acid in it, we couldn’t use it. However, we needed the negative ions in vinegar to bring together the polymers in plastic. We found that alkaline has a lot of negative ions. So, we looked for a vegetable that had a lot of alkaline and carbohydrates so that it will break down in water - through a process called hydrolysis - as well. This problem of having trillions of pieces of plastic in our ocean is important to fix because it is affecting not only ocean life, but us too. BPA in plastic has been linked to obesity, problems in reproduction, and cancer. It has also been found in 93% of people tested. Also, the plastic is made out of petroleum, so the production of plastic requires the burning of fossil fuels which contributes to climate change. The petroleum extraction sites also need to be cleared, so deforestation is also present, along with the fact that animals might mistake the petroleum after it is extracted, for water, which can further harm the environment.

Our project makes a difference in the survival and conservation of marine life and reduce ocean pollution (of plastic/trash). With biodegradable plastic, pollution in the ocean will decrease. At least 91% of plastic is not recycled… it eventually ends up in landfills or in the ocean. When it is in the ocean it can be mistaken for food by animals. When animals eat plastic, the plastic makes them feel full and they starve to death. So our goal is to reduce the deaths of the organisms in and out of the ocean. The plastic can also release harmful chemicals if it is sitting out in the sun too long and could poison the water, which can easily affect the plant life. We are trying to help the environment by making a plastic that is safe which can be safely decomposed even if they are not properly thrown away, as it degrades in 3 to 6 months. With the use of biodegradable plastic the plastic problems can cease. Also, we made a model of a machine that converts it into filament for 3D printers so that people can make whatever they want to make with this biodegradable plastic. Basically, the consumer would place the ready biodegradable plastic (from our unique formula) onto the stand, then the plastic would go through a series of conveyor belts with cold air blowing on them to make it dry. Then, it will reach a rotating pipe with an empty filament roll on it, and would get wound up to make filament for 3D printers. The pipe would be rotating by the use of a DC motor.


Mary Cotter

Allendale Columbia School, New York

Robots are the future. You see new videos from every consumer electronics trade show of new robots for your house, for disarming bombs or even personal companionship. They’re also incredibly interesting! Despite decades of research, there are still basic challenges that don’t have general solutions for all types of robots.

Inverse Kinematics (IK) is one of those challenges.  IK is the problem of determining joint angles for robotic arms in order to have the arm’s tip (called the end effector) reach a given point. It’s a basic problem in robotics, but it’s also an area that researchers still explore.

Analytical solutions to IK use exact geometric methods that lead to unique sets of joint angles for each point in space. These angle that can sometimes change suddenly for points that are close to each other. This can lead to lots of strain on the motors of the robot. Also, only very simple arms have a unique set of joint angles, so in addition to solving for the joint angles, an IK solver for complicated arms needs to provide some way of choosing between possible solutions. This selection process can also lead to big jumps, which look like glitches in our demo system but are actually correct solutions that cause big movements of certain joints.

Forwards And Backwards Reaching Inverse Kinematics (FABRIK) is a heuristic, iterative way of solving the IK problem. This was interesting to me because it uses current pose information to efficiently find an arm position and because it tried to find a “good enough” solution. FABRIK seemed like it could be applicable to a real robot arm.

Check out my video to learn if FABRIK really is a promising way to control simple robot arms!


Sneha Revanur

Quimby Oak Middle School, California

I have always been passionate about finding the solutions to some of humanity’s most paramount issues in a computational space; stories about the plight of women victimized by gender-based violence (GBV) around the world filled me with a drive to somehow leverage computer science to ameliorate their lives. I soon learned that little to no information about GBV exists, which is why it is so difficult for aid groups to effectively channel their efforts to tackle the issue. Since humans generate masses of data daily through social media, I thought that I could use Tweets, or posts on Twitter, to make conclusions about how we discuss and react to GBV. I went on to develop a computational algorithm that employs machine learning and natural language processing to independently label Tweets as GBV-relevant or not and classify them among four categories: Physical Violence, Sexual Violence, Harmful Practices, and Other. My model is able to do so with upward of 85% accuracy, and it can provide information about the types of users that discuss GBV, how they do so, and the geographical regions that most frequently discuss GBV.

Twitter has an accessible, secure API and offers a wide range of free streaming services. I set up and defined each of the four classes while also maintaining search criteria for each; Twitter’s services would allow me to only stream Tweets containing any of the preselected words. I streamed, preprocessed, formatted, and filtered the nearly 2,000,000 Tweets I collected. I continued to perform natural language processing on my Tweet set using the Natural Language Toolkit (NLTK). The frequency of all the terms that remained following the NLP stage was counted, and the terms with a frequency greater than 4 became unigrams. I formulated my bigram list by pairing up any two consecutive unigrams. I proceeded to label the 4,000 Tweets I had after the filtering stage while maintaining objectivity and consistency. I used the Naive Bayes classifier, which is easy to understand and implement, for machine learning. I conducted a comparative analysis of the two feature sets I was working with: the unigrams and bigrams generated from the NLP stage and the unigrams and bigrams generated from the search criteria I selected for streaming Tweets.

I found that the search criteria-based features gave me better accuracy and precision, with that feature set tending to perform better on average. Overall, I was able to develop a model whose classification accuracy was more than 85% to achieve my goal for the project. I was also able to mine information about the cities, states, and countries that most frequently discussed GBV; information about the specific classes that were most frequently discussed; and topics that were most frequently discussed.


Veda Vundela

Springhouse Middle School, Pennsylvania

Air pollution is one of the biggest threats that looms on our horizon. It’s been here for years and it’s only getting worse. The worse part about air pollution is how it can impact and affect people. Many people around the world can suffer from negative side effects due to air pollution. Millions of people of all age groups die every year because of air pollution and more suffer worse long-term side effects. Young children and babies are most affected by this because when they are out playing, they won’t be able to tell which areas have high pollution levels as easily as older adults can. The pollution levels can also lead to the death of children due to acute respiratory infections of the lungs and other airways.  Many of these people live in big cities and third world countries where they can’t tell which areas have a higher number of pollutants in the air because it all seems the same. But what if they could? My project uses a gas sensor to tell whether there are air pollutants in the air. Although there seem to already be other gas sensors out there, they only measure one or two gases, like oxygen, ozone, or hydrogen. They also cost hundreds of dollars and aren’t readily available for everyday usage. The gas sensor that I am using can detect many gases, such as smoke, methane, propane, and butane. My inspiration for this project comes from my friends and the area that I live in. Most of my friends suffer from asthma and the area that I live in is rather polluted with many factories everywhere. But there are places that aren’t as polluted or are too polluted. I started to wonder if there was a way I could figure out how much pollution there is in certain areas. I figured out that if I can use a small gas sensor and rig it to an Arduino Uno board and code it, I can help detect the levels of pollution in specific areas. The gas sensor lights up as it comes in contact with a specific gas. In the future, I can:

  • Make the project designs lighter and more portable, so that it is easier for young children, or even people of all ages to use it and keep it with them all day.
  • Add a LED display so that it can show the exact amount of gas at a specific point in time
  • Make it into a bracelet design so that people can always wear it everyday and it can just alert them about the pollutant levels when it increases
  • Add a buzzer and LED lights to my design so the person is better alerted at when there is a higher pollution level in the atmosphere.
  •  Lastly, I want to program an app for my device so a person can see what the pollution levels were like all the time and can tell which areas had a higher level or pollution and higher risk levels.

Sruthi Sankararaman

Rachel Carson Middle School, Virginia

The sound and images students running around in panic from the recent incidents in Parkland FL as well as horrific scenes in sandy hook elementary school are still etched in our memory.  Legislative option to control the fire arm violence in school premises is such a contentious issue even today. My aim is to use technology as the means and ways to make a small difference.  The core issue of my project is to how do we identify, track and if possible de-active the firearm. The recent advances in smart gun control technology provides a viable way to resolve this issue. 

The solution involves establishing a microchip embedded in the firearm device with Bluetooth transmitter and micro GPS tracker.  The Bluetooth will keep transmitting the unique Identification serial number of the firearm device that will help to identify the owner and details of the equipment.  The base monitoring station will keep reading this signal, identify the firearm, use the cloud services to do the background checks and create the alerts if the situation warrants.  As the Bluetooth allows two-way communication, the base station, can send the data communication to the fire arm equipment to activate or deactivate the device.

In my experiment, I simulated the firearm equipment using the Arduino board and mobile app as the base monitoring station. AWS cloud infrastructure was used for doing the background checks as well as for triggering the alerts.

The challenging aspect of this solution is twofold.  One related to developing the necessary legal framework to support implanting the microchip in the firearm equipment as well as getting the permission for tracking the firearm once it is in the public places.  The other challenge is related to allowing the technology infrastructure to support the background checks real time as well as making the firearms power operated.

Life is too precious to be lost to another gun related violence.  Every small step in the right direction counts.


Lara Scherer, Haruka Doi

Environmental Charter School at Frick Park, Pittsburgh Colfax K-8, Pennsylvania

The dictionary defines sexual harassment and assault as “harassment in a workplace, or other professional or social situation, involving the making of unwanted sexual advances or obscene remarks.” However sexual harassment/assault can happen anywhere and whenever. And anyone can hurt you from friends to complete random strangers. Sexual harassment could be traumatic and has been proven to have many negative impacts like depression, fear of the harasser, and not being social. There are already aids for victims of sexual harassment such as support groups, hotlines, the #metoo movement,  weapons to protect yourself, the emergency SOS on smartphones, and the assailants being put to custody. How can we prevent sexual harassment/assault before it happens? Our app attempts to fix that problem .

Our app called “Stoplight” is compatible with Apple Watches. The app on the watch looks like a stoplight, and includes a green, yellow, and red button. The green is for when you feel uncomfortable, but you are unsure whether you are imaging  a situation or not. It is labeled “Mild”. The yellow is for when you feel pretty uncomfortable. It is labeled “Severe”. And the red button is for when you feel extremely threatened. It is labeled “Emergency”. On the iphone, there are 2 pages on the first page there are 5 slots where you can fill on your contacts. On the second page, there are 3 sliders corresponding to the buttons on the apple watch. You can customize the timers. When you press one the buttons on the watch the timer that you filled in on the iphone app will run. When the timer runs out, there are different actions for the different buttons. On the green button it will send texts to the numbers of the contacts you filled in. On the yellow button it will call all contacts. And on the red button it will dial 911.

We used Xcode and Swift to program our app. We feel that “Stoplight” can help potential victims to not let it happen to them. Our mission is to prevent this serious issue.


Diya Satanur, Amulya Jayam, Anika Patchala

Oak Middle School, Annie Sullivan Middle School, Horace Mann Middle School, Massachusetts

Global health is a very big issue; one problem that affect our country is hunger. According to research by Feeding America, 1 in every 8 Americans struggle with hunger. About 41 million people in America are food insecure. Many grocery stores around the country try to help with this. They set up food drives and they also donate their extra food. Although grocery stores are finding ways to reach out to food drives, there is a better way. According to research, many grocery stores have their own systems to donate food. They find their own ways, but each grocery store is disconnected from the other. This is a very disorganized way, and grocery stores and food drives are often stuck with problems like having to transport the food. Our app, Food Donor can eliminate all these issues. It pairs people who want to donate food (Supermarkets) to the homeless shelters near them.

According to 2014 EPA Study “America  throws away more than 38 million tons of food every year. That is almost the weight of 104 Empire State Buildings”. The food we waste also costs out country $165 billion per year. We are also wasting  a huge amount of resources put into the production of these foods. When we throw away food, landfills are full and the organic waste decomposes releasing methane gas which is harmful.

In USA alone approximately 15 percent of the population live in homeless shelters. Majority of them include our senior citizens, women and children. Due to FDA regulations, there are myriad of labels on the food like “best by date”, “Used by date”, “Sell by date” , “Expiration date”. Supermarkets in America follow the guidelines of FDA and in this process lot of food gets wasted. Supermarket managers patiently explained the process as we took tours in the grocery stores. 

These items are called “Unsaleable items” and are disposed off the shelves. Some of them are thrown away while the remaining are taken by truck to the Product Reclamation Center. Walmart has its Product reclamation centers and Stop and Shop has its own reclamation center. These reclamation center sort out the items , throw away items that have broken in transit,  sell a portion of it at very low costs to the salvage store ( around 30%), throw away a portion which have expired (around 30%) and donate the rest to the food banks.

As we were enlightened about this process, we realized that food was getting wasted at many places. We want to help the homeless shelter  as well as reduce wastage of food at the supermarkets. To solve this problem, we built Food Donor.

Our app directly connects the supermarket with homeless shelters. The fruits and vegetables which are replaced because of the new delivery that the store received do not land up in waste bin but is used by many people who are in the shelter. Second, the cost of transportation these foods (fruits, vegetables, dairy, baked goods) to the Product Reclamation Centers is avoided as it is directly used by the homeless shelter. Last our app is saving earth as food is being consumed before it reaches the landfill. Food Donor is a win win for all us. Use our app and help the hungry.


Isabel Tucker, Amanda Gilpin

Jefferson Middle School, Tennessee

Each year, 150,000 people could be dying because first aid is not known enough worldwide. That means that because of the lack of first aid, the same amount of people die to those who die because of cancer. Suffocations claim about 2,500 lives and heart attacks kill about 29,000 people each year. In a campaign launched by St. John Ambulance to get more people learn first aid skills, a group of more than 2,000 people found that 59 percent of people would not be confident in saving someone’s life, and 24 percent of people would do nothing if they saw someone struggling and just hope that someone else knows first aid or an ambulance will arrive. It also revealed that 44 percent of U.S. adults do not have first aid kits and 48 percent lack emergency supplies for catastrophic events. 60 kids die at home each year because of accidents, but a quarter of parents still have no first aid supplies. A recent study shows that only 31 percent of parents actually knew how to help their child and were confident during an emergency. On the other hand, 38 percent of parents were positive that their child was going to die, and 5 percent were worried that their child would be left injured. It has been stated that 23 percent of parents have actually experienced their child choking, 14 percent of children have reported to have a seizure, and 1 in 20 have an allergic reaction. It is sad to say that despite these facts, 6 in 10 homes still do not have first aid kits. It also revealed that 55 percent of Americans rely on local authorities to come to their rescue when disaster strikes. That is why we decided to choose the theme “Global Health”.

We wanted to make a first aid box that could provide help to the user so that people could provide first aid even if they are not confident or unsure of what to do. The Smart Medical Box has two components; the app and the box. The app provides a list of injuries with information on each one, and if you click on the injury, an LED light inside the box will light up telling what medical supplies are needed and tells you what medical procedures you need to do. The box has two drawers that have medical supplies in them. In each drawer there are different sections, each with an LED light that glows when you click on an injury in the app.

For our coding we used an Arduino with 3 pre-existing codes that we modified to fit our needs, a Bluetooth module to connect to the app, a power bank to supply power to everything, and a heart rate monitor. As you can see, the Smart Medical Box can be very effective in multiple situations, and can help many people everywhere


Deepika Kannan

Sanford Middle School, Florida

Every year, an increased number of people die, due to the lack of steady communication. Li-fi, or light fidelity, is an upcoming form of communication which uses visible light to transfer data. LED bulbs, whose light intensity can be changed very rapidly, are used to transmit data to a photodetector. The photodetector receives the light signals and converts the data into a “stream-able format.” Compared to wi-fi, li-fi has a bandwidth (range of frequencies) which is almost 1000 times more, this enabling more rapid transfer to data. Although li-fi cannot pass through walls, this makes its data transfer more secure. Li-fi is also said to hold fewer health hazards relative to radio waves. It won’t interfere with radio waves produced by medical equipment like MRI scanners. If li-fi is incorporated outdoors, even our street lights can provide us with high-speed internet. In the future, it is thought that li-fi communication between LED headlights on one car to another can help minimize accidents. LEDs can thus be used in a dual function for illumination as well as communication.

My main goal of my project was to determine the feasibility of incorporating li-fi in home, office, and outdoor settings. I first designed an LED based circuit capable of transferring audio signals. I used a commercial software to create multiple sound samples of varying frequencies and volumes. I measured sound intensity as a function of distance between the transmitter and receiver, using Bluetooth as a baseline. Subsequently, I performed similar experiments using Li-fi. Next, I repeated these trials across various barriers, including paper, glass, and fog/pollution. However, there were no significant differences between data transfer between air and glass barriers.

From my project, I learned that li-fi can be incorporated in home, office, hospital, and outdoor settings. Interior decor for homes or offices made of glass will not hinder the quality of data transmission. On the other hand, paper offers privacy solutions against hacking. The quality of data transmission was not affected by the barriers intended to replicate fog/pollution, so li-fi is also feasible in external environments.