Cornell Bowers College of Computing and Information Science

BOOM 2019 Projects

Thanks to all our BOOM 2019 project teams!

Cornell University Autonomous Underwater Vehicle

Tal Moriah, Anthony Viego, Kevin Guo, Cuyler Crandall

The Cornell University Autonomous Underwater Vehicle (CUAUV) team designs and builds AUVs for competition and research. The team and its vehicle have received acclaim and support from industry professionals as a result of its performance and long-standing tradition of excellence.

JASPR: Just A Stupidly Persistent Robot

Switzer Ian

Provided with little more than a camera and a finger, Jaspr uses the combination of Computer Vision, Convolutional Neural Networks, and Q Learning to train himself to reach humanly impossible scores on popular smartphone games such as Flappy Bird, Jetpack Joyride, or Hill Climb Racing.

Let's Hear It: Female Empowerment Platform

Maya Frai

Let's Hear It is a female empowerment platform and publication that features stories of leading women in diverse industries. LHI has interviewed women such as Kara Goldin (CEO and Founder of Hint Water),  Cheryl Yeoh Sew Hoy (Founder of the #MovingForward Campaign and MaGIC Malaysia), Aanchal Gupta (Director of Security at Facebook). Through online campaigns, op-eds, interviews, and contributory pieces, LHI aims to inspire, advise and empower the next generation of women.

CUAir (Cornell University Unmanned Air Systems)

Anna Fang, Sitar Harel, Chaska Yamane, Ben Sternklar-Davis, Jamie Gray

CUAir is an engineering student project team that designs, builds, and operates a custom search-and-rescue unmanned aerial system. CUAir is the top fixed-wing aircraft team in the world, competing in the annual AUVSI SUAS competition and soon entering the Australian Medical Express Challenge in 2020. Our team consists of 7 subteams and over 60 members, spanning majors such as Computer Science, Mechanical Engineering, Electrical Engineering, Applied and Engineering Physics, and more.

Blockchain Enabled Agro-Economic Credit Scoring

Shiang-Wan Chin, Kriezi Foteini, Yizhuo Yu, Shukla Mukul, Rylander Haley, and Xueyuan Wang

We are using blockchain enabled digital valuation based on soil and other agroeconomic indicators to enable credit scoring for farmers. We use drones and satellite imagery as a service to feed into our machine learning model to calculate credit scores.

Computational Analysis of the French Revolution

Isabelle De Brabanter

For the academic year of 2018-2019, I have been working on a senior honors thesis under the guidance of PhD candidate Laure Thompson and Professor David Mimno. We have been building off of the work done by Barron, DeDeo, et al in their project "Individuals, Institutions, and Innovation in the

Debates of the French Revolution". By scraping and cleaning XML files provided by Stanford University's French Revolution Digital Archive, we are able to run extensive text mining computational analysis.

Apomden

Samuel Opoku-Agyemang, Sena Katako, Dennis Delasi Nyanyo

Apomden is a health technology application that efficiently handles bed management and availability for health facilities in developing countries with low internet connections. We also aim at streamlining the transfer of patients between hospitals hence, reducing the risk of referring or turning a patient away without securing a bed in another facility for them.

Cornell Mars Rover

Andrew Arroyo, Daniel Hirsch, Tanmay Bansal, Leigh Harrison, Jane Du

Cornell Mars Rover is a student-run interdisciplinary project team composed of 40 members. Each year we build a rover to compete in Utah against other university teams from around the world. The rover needs to be capable of performing on board chemical analyses to test for signs of life, traversing rough terrain, autonomously navigating a course and detecting visual markers, and achieving precision control with the robotic arm to complete human-like tasks.

Flux

Neha Rao, Andrew Xiao, Matthew Coufal, Kaushik Ravikumar

Flux is a mobile application that tracks and reports the traffic flow of on-campus dining halls and cafes in real time. Flux aims to provide students real-time data to help them plan ahead and find the best places to eat at the best times. Available on the iOS App Store and Google Play.w Xiao, Matthew Coufal, Kaushik Ravikumar

Comment Toxicity

Debasmita Bhattacharya, Ruchika Dongre, Nikhil Saggi

We are building a multi-headed model that is capable of detecting and classifying different types of toxicity, such as threats, obscenity, insults, and identity-based hate, in online comments. We use supervised learning to accomplish our classifications and compare the following methods: Gaussian, Naive Bayes, K-Nearest Neighbours, Random Forest, Decision Tree, and Logistic Regression. Finally, we examine the differences between different feature extraction tools.

Cow Mastitis Prevention System

Sonia Appasamy, Nehal Rawat

Mastitis is a prevalent disease amongst dairy cattle and causes inflammation in their mammary glands and udder tissue. Not only is this disease potentially fatal for cows, but it also causes many issues for farmers, as they take on large treatment costs and see a decrease in milk production. We hope the system we design will help to predict mastitis and the type of pathogen present in a more timely fashion so as to minimize the negative outcomes of the disease.

Cow Pregnancy Prediction

Bhatnagar Arshi, Tian Ren, Xueman Liang

We aim to predict the probability of pregnancy based on milk yield, milk components, body temperature, environmental temperature, body condition score, cow features in a scalable fashion using Microsoft Azure. Our project will allow for information to come in from sensors and appropriately be processed.

Real-time City-Bike Stocking via Ad-hoc Incentives

Swathi Chakrapani, Mindy Lou, Christine Shen

Many cities have adopted city-wide bike share programs where users can rent bicycles from centrally placed docking stations and drop them off at another station near their final destination. Over the course of the day, these bikes may be sub-optimally displaced

Mixtape - Playlist Collaboration

Austin Astorga

Mixtape allows friends with different streaming services (Spotify & Apple Music) the ability to create collaborative playlists. The playlists will sync back to the user's specific music app of choice. Friends now will have the ability to share songs and create playlists with people using a different streaming service.

Fake News Detection

Linnea May, Kathy Byun, Danny Yang, Eric Sun, William Bekerman

With the rise of the threat of fake news, more and more approaches are being taken to try and identify it. Our project uses Deep Learning and Natural Language Processing to fact check headlines as an extension of the 2017 Fake News Challenge, with the added goal of making results more interpretable to the user. 

Robotic Rubik's Cube Solver

Eric Berg, Yangyi Hao

My partner and I have created an automated robotic Rubik's Cube Solver!! It uses a Raspberry Pi camera to scan the cube and then is able to implement a solving algorithm using mechanical claws powered by stepper motors and solenoids. The entire chassis and electronics system was customized designed by us!

Cloud and Edge Computing for Smart Herd Management

Xinzhe Yang, Yuntian Lan, Sixian Yi

By applying cloud computing technologies, our project attempts to predict cows' calving date and time using data on its body temperature and physical activity using data collected and transmitted from the SmaXtech sensors. With success, this project will be able to automate the process of birth prediction and give useful advice and reminders about calving to get the farm prepared, greatly improving the accuracy and efficiency of the whole process.

Indago

Joanna Davis, Sungjae Ahn, Arun Chadhuri, Hannah Mahanti, Kaavian Shariati

It takes an average of 2 years for a research group to file an application to get their drug approved for clinical trials, which keeps life-saving medicine away from someone in need. Indago is an interface that enables research groups to complete the forms that required to submit to the FDA to get clinical trial approval. We aim to maximize efficiency by streamlining and centralizing the process.

Poloma Analytics

Jack Pertschuk, Cole Theines

Poloma Analytics is dedicated to improving the efficiency of bringing novel therapeutic compounds (drugs) to market. We utilize unprecedented amounts of unstructured (text) and structured (database) data to predict the real-time likelihood of a compound’s ultimate regulatory approval.

INSPIRE: Insightful Spotify Recommendations

Max Chen, James Chen, Shalin Mehta, Brandon Truong, Sam Fuchs

Spotify currently provides song recommendations, but provides no indication of how the recommendation is made. Other streaming platforms like Netflix state which show invoked a recommendation. INSPIRE is an application that provides song recommendations with insightful comparisons.

Carriage

Max Chu, Jessica Hong

Paratransit service in the United States is powered by technology stuck in the 1990s. Carriage provides paratransit with modern ridesharing technologies -- an automatic ride scheduler, companion driver/rider web applications, and service management for dispatchers.

uniVRsity

Nicolas Barone, Chad Fong, Rigolosio Domenic, Mutolo Jeremie

uniVRsity is revolutionizing online learning through Virtual Reality(VR) by creating the first VR STEM tutoring app. No longer will students suffer from the distractions from what's around them and the boredom that comes with online learning. Through VR’s immersion and interactive experiences, students will be focused and actually enjoy learning.

cue: Events at Cornell, just for you

Amanda Ong, Lisa Barbera, Jill Wu, Jagger Brulato

Cue is the premier app to discover all the events at Cornell. With the goal of creating a more connected, engaged community, cue helps students discover new activities to try and find new organizations to join. Students can follow the clubs they care about and stay up to date on the events happening on campus. Cue is being developed for iOS and Android, coming soon.

Eatery

Conner Swenberg, Zain Khoja, Cecilia Lu, Catherine Zhang

With over 6,000 students using it every month, Eatery enables students to browse menus and discover places to eat on Cornell’s campus. It enhances the dining experience at Cornell with features such as checking meal swipes and dining money balances and soon-to-be favoriting dishes, providing the crowdedness of eateries, and collegetown restaurants. Eatery is available on both iOS and Android platforms.

Lifestyle Politics on Facebook

Juliana Hong, Hannah Lee

The rapid growth of social media data allows computational social science researchers to now examine human behavior at an unprecedented scale. To better understand the political partisanship of different lifestyles (e.g., enjoying lattes vs. hunting), we constructed a unique dataset of 300,000 Facebook interests. We identified polarized classes of interests and "cultural bridges", as well as examined the driving forces behind group polarization by decomposing their conditional distributions.

Cornell Autonomous Bicycle

Aditya Narayanan, Sebastian Colom, Glus Daniel, Mar Jason

The CU Autonomous Bike project team is a multi-disciplinary team of  mechanical engineering, electrical engineering, and computer science students working together to build a robotic bicycle with the ability to balance itself  the same way a person balances on a bicycle: using steering manipulation.

LIBRA

Li Chai, Zeya Peng, Andrea Gonzalez-Del-Valle, Yue Sun

LIBRA is a wearable that helps victims of domestic violence to collect evidence and empowers them to leave the toxic situation.

RELI Delivery System and Incinerator

Victoria Tan, Jacob Mathai, Evan Cheng, Ana Elhom, Sham Hamidian

The Integrative Design Subteam of the Engineering World Health Project Team is collaborating with PATH on a reusable, electricity-free, low-cost infusion pump, a device that is able to deliver various types of fluids at controlled flow rates to patients. The subteam is also working on a solar-powered menstrual waste incinerator with Barefoot College so women in rural India can dispose of their menstrual hygiene products in a comfortable, convenient, and environmentally-friendly way.

Cowzure

Michelle Ip, Zack Brody, Janice Chan

Cowzure is an integrated smart farm solution that predicts whether a pregnant cow will calve within a certain time frame, and alerts a farmer via Facebook Messenger so they can adequately prepare for the cow's birth. Our project uses Microsoft Azure to ingest sensor data from cows and analyze it in real-time to give the most accurate predictions possible. Cowzure also provides an interactive dashboard to monitor each cow’s overall health metrics and see aggregated data for all cows on a farm.

Loxodonta

Camryn Gray, Kerri Diamond, Seth Brunell, Priya Malla, Anna Huang

Our project, Loxodonta, is an action-packed video game where African elephant L427 is being held prisoner on an alien research facility and must fight to find her way home to Earth. Players utilize L427's stolen hoverboard and taro leaf to dash into enemies, push them into pits, and navigate through rooms in search of escape. By exploring and finding the access key in each level to unlock the next, players bring L427 closer and closer to sweet freedom.

Kamachi

Annie Barnett, Janie Walter, Eddie Hew, Cindy Xu, Max Gaeta

Kamachi is a city-building and exploration game. Kamachi, a baby city-turtle, has just wiggled out of its egg and must undertake a dangerous pilgrimage to reach its true home: the ocean. Help Kamachi by cultivating a city atop its back, while traversing through mysterious, folk-tale-esque landscapes. Along the way, collect resources to feed the villagers and manage its grounds, but be aware of unforeseen events that could bring Kamachi’s journey to an abrupt end.

Flourish

Jeffrey Yao, Michael Xing, Wenjia Zhang, Demi Chang, Samuel Sorenson

In the world of Flourish, Homo sapiens have abandoned an exhausted Terra and life is beginning to reclaim its place on the surface. In this game, players take on the role of a plant attempting to grow in an environment with a scarcity of resources and a plethora of competition. Using biomass, water, and sunlight, the players must spread roots and grow towards the sun. At the same time, players must keep a watch on rival plants while balancing their own resources to prevent decay.

Amaris: Realm of dreams

Tracy Goldman, Matthew Shih, Mina Huh, Niara Boykin, Ikra Monjur

Stuck in a cold, lonely reality, an imaginative child named Amaris eagerly fulfills her fantasies of power by night in her moon-powered dreamscape until a dark evil invades, shattering the moon into several shards. Empowered special abilities obtained from entering dream fountains to grant her super speed, limited flight, and incorporeality, help Amaris collect all of the different moon shards and restore life to the moon while maintaining her dream state.

Ezra Box

Yi Hsin Wei, Sophia Deng, Andrei Kozyrev, Rishabh Sarup, Ally Yuan

Ezra Box is a peer to peer storage platform that connects Cornell students who need space with Cornell community members who have space. Our pilot last summer catered to the storage needs of +150 students on Cornell's campus. Ezra Box is 50% cheaper and more flexible than any other storage alternative available to Cornell students. We're a project run by and for Cornell students to improve the local community and bridge a cap in the current market.

GiveGrow

Varun Iyengar, Boonakij Palipatana, Emily Chan, Jennifer Lee

5 dollars can provide 50 meals to people in need. A longstanding misconception about charity is that only large donations can make a difference. GiveGrow shows users exactly how far their money can go and makes donating more accessible and rewarding. Users designate a small percentage of their weekly income for donation and watch as it grows over time. The app shows users how much that money can buy for charities involved with causes such as conservation, medical research, and disaster relief.

AnInverstor: a News-Based AI algo trading strategy

Lily Liu, Betsy Fu, Jonathan Ni

“Amazon Pulls Out of NYC Headquarters...” We have been reading news like this every day and markets swing wildly on even the smallest of news. While it is very natural for humans to read a headline and make rational predictions about the stock market, an AI trading algorithm makes a great effort to “understand” the words and to “decide” when to buy or sell.  We propose a news-based AI trading strategy that utilizes NLP and ML techniques to predict the stock trends and make unbiased decisions.

Starstruck

Arielle Goldberg, Shan Parikh, Joy Zhang, Sheri Guo, Hye Won Kim

With an empty galaxy to call home, Nova and David seek to fill it with light. Traveling from planet to planet throughout the universe, the duo surrounds stars with the rope that connects them to collect and fill their galaxy with constellations.

Playing as both characters, the player must strategically jump from planet to planet and anchor appropriately in order to successfully collect all of the stars from the level in the puzzle adventure game Starstruck.

Clerical Error

Elaine Hwang, Jessie Yee, Jane Chen, Matt Davis, Jackson Rong

Clerical Error is a action puzzle platform game based in hell. Sometimes, even gods of death make mistakes; prematurely trapped in the Underworld due to a clerical error, Cid’s soul isn’t properly detached from their body as it should be. Use Cid’s newfound ability to separate their soul and body to trick demons and ultimately escape this bureaucratic hellscape.

Sync-up

Calvin Sowah, Jannie Li, Bella York, Bryan Kamau, Emily Leng

Sync-up is a comprehensive recruitment solution designed to simplify the application process for internships and for full-time positions for students in the Ivy League and for recruiters. It combines cloud computing, artificial intelligence and NLP to generate direct matches between students and specific teams at Fortune 500 companies based on the interests, values and skills of students.

YoyoBoy

Everett White, Kexin Zheng, Chris Oh, Ariel Lin, Daisy Dai

While exploring an abandoned factory, a young boy ends up falling through the decaying floor into the bowels of the building. Armed with just his wits and a yo-yo, the boy must make his way up from the factory depths by using his yo-yo to swing, grapple and slide through each challenge. This puzzle platformer will have players controlling the boy and his yo-yo in order to guide him through each level and solve puzzles so he can get back to the surface.

Psafe

Calvin Sowah, Clive Duncan

Reaching out to mental health professionals can be a struggle for students especially when those professionals do not relate to the students culturally and socially. This can be particularly difficult for students that are new to campus life. Psafe is a mental health solutions platform focused on addressing the diverse mental health challenges of students (particularly, minorities) by matching them to psychiatrists based on the students' demographic and socio-cultural preferences.

Avicaching: Reducing Bias in Citizen Science

Anmol Kabra

Citizen science programs, instrumental for sustainability projects and scientific discovery, often suffer from sampling bias due to citizens' biases in collecting data. For eBird, a bird observation program, Avicaching learns citizens' behavior to optimally distribute incentives, which motivate citizens to visit unattractive bird-viewing locations. We achieve 800x speedup compared to the previous state of the art while performing comparably, thus assisting citizen science to scale efficiently.

Strata with Lease

Junkai Liu

Strata is a cross-media file system on a single node. Strata implements multithreading file access control. However, it lacks concurrent file access control across processes. In this project, we fill this gap by implementing the lease. Our design is optimized toward write-intensive workload. Our implementation provides the necessary infrastructure to make Strata become distributed file system.

Generating Piano Music with Self-Attention

Ziwei Gu

Can machine learning play a role in the creative process? Do computers have an advantage in generating more complex sequences? How to strike a balance between novelty and rules/patterns in AI systems? This project explores these questions and generates piano music with long-term coherence, creating more than 8 consistent phrases (around 40s) out of the given motif.

Pod Acai

Caley Drooff, Keivan Shahida, Yanlam Ko, Drew Dunne, Manish Saha

We are creating an app for food pre-ordering aimed at helping driving new business to local eateries. We have begun with a partnership with a local Cafe called Chatty Cathy to test the initial design, structure and user flow of such an application type.