Cornell CIS

BOOM 2018 Projects

Thanks to all of our BOOM 2018 project teams!

CUAir

Shirley Zanifur Kabir, Troy Joseph, Nikita Ermoshkin, Sitar Harel, Chaska Yamane

CUAir is an interdisciplinary student team that designs, builds and tests a custom search and rescue unmanned aerial system. We are the world champions of an international unmanned aerial vehicle competition.

Deaf Dance

Philina Chen, Wenchang Yang, Yipeng Zhang, Hanchun Shao, Yicheng Zhu

Over 360 million people worldwide have some degree of hearing loss. Dance is widely recommended for people with hearing loss as it extends their social interaction and enables greater self-expression. However, limited self-consciousness, access to specialized dance studios and learning resources impede their passion for dance. Here we introduce Deaf Dance, a wearable device and mobile app that empowers deaf and hard-of-hearing people with engaging multisensorial dance learning experiences and extended networking opportunities. The wearable device vibrates in sync with dance rhythms and the mobile app features accessible dance tutorials, music visualization and dance partner finder. We implemented an iterative human-centered design lifecycle, including user research, ideation, information architecture, wireframing, prototyping, interaction design, and user testing. The user research included a literature review, user interviews with 13 amateurs, contextual inquiries with 5 dance professionals, and the observation of a 2-hour rehearsal at a specialized deaf dance company.  Deaf Dance aimed to provide engaging self-taught dance learning experiences to elevate deaf people’s self-expression and social interaction with larger communities. So far we finished the hi-fi prototypes for the mobile app and initial design of the wearable device.

Robomixer

Lola Legrand, Sam Hamburger, Gillian Boehringer

Robomixer serves delicious, precise cocktails, requiring virtually no effort. Watch the Robomixer automatically deliver quality drinks as you enjoy the lightshow.

Felongeist

Nicolas Kuhn de Chizelle, Catherine Zhou, Nicole Tan, Stefan Joseph, Vincent Imbimbo

A fun puzzle game where players help an injured criminal escape a hospital by controlling his spirit to possess and manipulate hospital staff.

Roboto

An Tran, Annie Cheng, Scott Wu

Roboto is an interactive and expressive dancing robot companion. With an approachable and adorable design, Roboto will dance to music in a principled fashion and interacts with the user in an expressive manner.

Recast - Find & Share Podcasts

Natasha Kristin Armbrust, Mindy Lou, Eileen Dai, Madeline Ugarte, Kevin Greer

Recast is an iOS app made by Cornell App Development where friends can discover and share podcasts effortlessly. As a user, you'll experience both a fully functioning podcast player while simultaneously discovering trending episodes and series that are customized to your own interests. Through a feature called ‘Recast’, you can share podcasts with friends and connect with a community of podcast listeners. A stellar recommendation feature and fresh user interface ensure that you are always exposed to original podcasts that expand your knowledge and creativity! Recast’s vast range of features allows for an optimized and personalized listening experience: - Discover top series & episodes through recommendations and topics - Search for series to subscribe to, episodes to listen to, and users to follow - Recast episodes to share podcasts publically - Share episodes personally with friends - Specialized news feed which shows recommended episodes to subscribe to, recommendations from friends, and series friends have subscribed to - Save episodes to listen for later - Listen to podcasts at .5x - 2x speed - Create a personalized profile - Follow your Facebook friends - Facebook & google login

CU AUV

Nicholas Akrawi, Zander Bolgar, Laura Lee, Artina Maloki

CUAUV is a completely undergraduate project team composed of 48 students, spanning 6 majors, that competes in AUVSI’s international Robosub competition. Each year, we design, build, program, and test two completely new autonomous submarines. In July, we ship our subs down to San Diego, put it in a 200'x300' Navy testing pool, press the green button, and cheer it on as it navigates a complicated underwater obstacle course with no human intervention whatsoever. Two years ago, CUAUV was the first team to develop two vehicles in parallel, forever changing the way teams compete. Last year was focused on building upon our success with two submarines, and our hard work was realized in July when Artemis and Apollo took home the team's seventh first place win at the 20th Annual RoboSub. This year we are hoping to continue pushing the limit of AUV technology with our most compact and capable vehicles to date, Castor and Pollux.

QuACC (Question Answering for Cornell Courses)

Kenta Takatsu, Yuji Akimoto, Chetan Velivela

QuACC (Question Answering for Cornell Courses) is a machine intelligence system that improves the learning experience of Cornell students. This system is capable of synthesizing information from raw text resources, such as textbooks and syllabi, and responding to questions about those resources. We trained this system with Wikipedia articles from the SQuAD dataset (Stanford Question Answering Dataset). In the future, we will experiment if the model can answer Piazza questions with the knowledge from textbooks. We hope to integrate our system with online platforms like Piazza to help Cornell students by automatically answering simple questions. This project was supervised by Professor Thorsten Joachims, and implemented by 3 engineering students from the Cornell Data Science research team..

Duplicate Question Identification (DuQI)

Zhao Shen, Brandon Kates, Arnav Ghosh

Our goal is to create a system capable of detecting duplicate questions on Q&A platforms such as Quora and Piazza. We expect our approach to help centralize the available knowledge on a single question/issue and direct users with questions that have already been answered to the appropriate resource, so that people spend less time searching and more time learning. From support vector machines to recurrent neural networks, we will use a variety of machine learning models on the Quora question pairs dataset. We will find out which predictive techniques work the best and why they perform so well, hoping to eventually apply our findings to develop a highly accurate yet generalizable model for use on the popular classroom Q&A platform Piazza. This project was supervised by Professor Thorsten Joachims, and implemented by 3 engineering students from the Cornell Data Science research team.

CU Reviews

Brooke Docherty, Julian Londono, Jessica Chen, Nihar Sidhu, Divyansha Sehgal

CU Reviews was developed as a way for students across the university to get advice about courses offered on campus to help them better choose what classes to take. The current application, hosted at CUreviews.org, allows anyone to search for a class at Cornell, leave an anonymous review, and view approved reviews and aggregate review data. Currently, this data includes the estimated quality, difficulty and median grade of a class, as determined by the average input of previous reviews. In the coming weeks, we plan to alter the format of reviews to additionally provide information about the workload required by the class, as well as how highly recommended it is by other students. In the months since the release of the current version of our application, we have received over 300 reviews from Cornell students and have over 1,500 all-time users. As our product continues to evolve, we plan to market it more extensively to encourage students to contribute reviews about their class experiences. After much user research, we have confirmed that many students lack relationships with older students who can provide them with this kind of honest advice. The hope is that our tool will help solve this problem and serve as a valuable resource for students who want to know more about classes they are considering enrolling in.

Ithaca Transit - Live Tracking

Matt Barker, Monica Ong, Austin Astorga, Mihir Chauhan

Introducing Ithaca Transit, a new end-to-end navigation service for built for the TCAT bus service. A free and open-source app, Ithaca Transit offers a diverse range of features in beautiful, clean interface to help get you where you need to go.

Live Bus Tracking

Want to know is the bus is running late (again) or where it’s even at? We’ve got you covered. View actual bus locations on the map and see expected delay times, all in one easy-to-use interface. We use the latest dynamic transit data directly from TCAT to ensure the most up-to-date information.

Search Anywhere

Ithaca Transit integrates with Google Places to allow you to search bus routes to any destination in the country. Search up Chipotle or Waffle Frolic and let the app take care of the rest, including accurate walking directions!

Your Favorites

Easily bookmark your favorite bus stops and destinations for one tap access to routes. Blazing!

Made by Cornell AppDev

Cornell AppDev is an engineering project team at Cornell University dedicated to designing and developing mobile applications. We were founded in 2014 and have since released apps for Cornell and beyond, from Eatery and Big Red Shuttle to Pollo and Recast. Our goal is to produce apps that benefit the Cornell community and the local Ithaca area as well as promote open-source development with the community. We have a diverse team of software engineers and product designers that collaborate to create apps from an idea to a reality. Cornell AppDev also aims to foster innovation and learning through training courses, campus initiatives, and collaborative research and development. For more information, visit our website and follow us on Instagram.

EmoteNote (formerly Cœur)

Jerica Huang

Emotions are data. When we label our emotions accurately, we are more able to discern the precise cause of our feelings. The readiness potential in our brain is activated, allowing us to discern the right steps for us. EmoteNote is a journaling app that uses natural language processing, sentiment analysis, machine learning, and techniques from coaching conversations to gently encourage writers to label and understand their emotions more accurately through annotations and prompts, rewards them when they answer prompts, monitors their emotions over time, and provides insights about values. Through such encouragement to accept and understand all emotions—even the messy, difficult ones—we realize the cornerstone to resilience, thriving, and true, authentic happiness.

In-Domain Cross-Writer Sentiment Classification on Movie

Xinzhe Yang, Xuwen Shen, Katie Yang

With the rise of the information age, sentiment analysis on texts has been a crucial Natural Language Processing task that many researchers have put effort on. Recent researches on sentiment analysis have obtained great results using various methods such as Naive Bayes, SVM, Recurrent Neural Networks and etc. More recently, cross-domain sentiment classification that applies the model trained on texts in one specific domain, for instance, reviews to the texts in another domain, for instance, tweets has been a topic of interest because datasets without standard labels can be classified using a pretrained model, where normal sentiment classification models give significantly worse results. [3] Popular methods for cross-doamin sentiment classification include Sentiment Sensitive Thesaurus [1], Stacked Denoising Auto-Encoders [2], Spectral Feature Alignment [4] and etc. However, little research has been done on "In-Domain" but "Cross-Writer" sentiment classifi- cation. In particular, for this task, the model is trained on the texts written a specific person and tested on the texts written by another person but in the same domain. For instance, apply the model trained on texts by a movie reviewer to the texts by another. In-Domain Cross-Writer sentiment classification is meaningful because we different people tend to have different writing styles, especially for highly subjective texts like movie reviews. While it is clearly different from normal sentiment classification tasks, it is also worth being separated from Cross-Domain tasks in that the datasets still share a common topic and the possible feature variables are reduced such that the focus will be on features like writing styles. Our goal is to conduct an empirical investigation on the performance current methods on this In-Domain Cross-Writer task and gain insights from the results to find the bottleneck and propose a new method.

Pollo - Polling Made Easy

Jonathan Y Lee, Jehron Petty, Jason Wu, Jack Schluger, Kevin Chan

Ask questions and take live polls with the click of a button with our iOS mobile app, Pollo. Through Pollo's simple, easy to use minimal interface, you are able to ask your friends, team, students, or audience questions and receive live feedback to make faster decision and better gauge audience sentiment.

ExploRentals

Jim Li, Kenta Takatsu

ExploRentals is a new way to find the best rental options for newcomers to the neighborhood. Imagine you are an incoming Cornell graduate student and are hoping to find an apartment in Ithaca that fits you won’t regret. Without knowing much about the local community, she can now go on ExploRentals to find a neighborhood that’s low crime rate, convenient for grocery shopping, and easy to get to Cornell campus. That’s thanks to ExploRentals’ “Local Info” feature, which allows users to overlay demographics and living information on top of the rental results on a map. Users can also star rentals that fit their criteria as they explore through the different map layers from “Local Info”, and review the starred results in a “Shopping Cart” at a later time. Furthermore, the starred results are intelligently ranked in the “Shopping Cart” based on the search criteria selected.

Cornell Mars Rover

James Haber, Andrew Showers, Eliot Huang, Tyler Wang

The Cornell Mars Rover project team designs an innovative Mars rover to compete in the University Rover Challenge. We are an interdisciplinary student-run team that brings together talented minds from engineering, science, and business. Together, we foster creativity through the development of our competitive design.

Research Connect

Aram Chouldjian Baghdassarian, Rachel Nash

The abundance of undergraduate research opportunities is one of the great things about Cornell, yet unfortunately the process of finding these opportunities is still very unstructured. One must knock on doors, send countless emails to professors, and make do with obsolete web pages to seek research opportunities. Research Connect aims to bridge this gap by providing a structured platform where students can find opportunities, and research labs can find the students they are looking for. STUDENTS 1. Easily fill out a profile with your credentials, interests, and resume. 2. Find listings that fit your interests and qualifications. 3. Apply and get in touch with researchers and labs in your field! LAB ADMINISTRATORS 1. Customize listings to reach hundreds of interested students. 2. Easily sort through applications and get in touch with qualified students. 3. Get the reliable research assistance you need in your lab!

Spotif-AI: Social Graph Music Recommendation

Abhimanyu Amit Gupta, Andrew Peterson

Spotif-AI is a music recommender which is trained on a user's social graph to best produce a playlist that is tailored to his/her listening tastes. A user logs into the Spotif-AI website and connects to their Spotify account. The user will then select up to 5 friends they wish to have influence a new playlist. Spotif-AI trains its neural networks on the listening habits and history of each person, producing a 20 song playlist which is tailored to the original user's preferences. The original user can further provide feedback to the application, citing a thumbs up or down reaction to each song. The algorithm then retrains based on this feedback to produce an optimal playlist for the user to enjoy, and can be sent to Spotify for in application listening.

Banana Maven

Amy Lin, Ziwei Chen

Banana Maven is an artificially intelligent solver that can play Bananagrams and solve the game faster than its human opponents. Bananagrams, a successor to the popular word game called Scrabble, is a tabletop game in which one arranges one's tiles into a grid of connected words faster than one's opponents. The objective of the game is to be the first to complete a word grid after the pool of tiles has been exhausted. Banana Maven generates and selects the best words to place on the board while arranging and rearranging words into ideal formations. The solver is built upon the application of game trees, depth-first search, randomized algorithms, dictionary API calls, and hill climbing. Hence, the resulting AI prunes its game tree efficiently without executing costly backwards scans of alternate board states. We present a demo along with architectural diagrams and explanations, reasons for not choosing alternative prototyped considerations, challenges, and next steps.

UpLift: Engage Community with Cornell Fitness Resources

Tina He, Jesse Prague, Keivan Shahida, Young Kim, William Bai

A team of 2 designers, 2 iOS developers, and 3 backend developers came together with a shared goal of making Cornell's fitness experience better and more engaging. Uplift provides students with information on all fitness centers, classes, and resources throughout Cornell. Feel like doing some cardio? Just search for cardio classes and we will show you when and where they are happening. Get information on gyms - from current occupancy to a detailed list of equipment. Invite your friends and share classes! Fitness is being developed using Swift.

WikInsite

Xinqi Lyu, Linnea May, Ziwei Gu, Nikhil Saggi, Jim Z. Li

The problem with learning nowadays is not the lack of resources but the information overload. As is often the case in studying, one may encounter a new topic and want to learn more about it. WikInsite investigates the paragraphs and links from Wikipedia pages to establish a useful topic expansion for our users with a chosen topic. We use the chosen topic as the basis and employ a few algorithms to find some most relevant keywords to it. The user can then further explore on our website, which has graph of the relationship between the topics and the actual links to the Wikipedia pages. As a result, the user can quickly learn more about the new topic he/she is interested in, without wasting too much time filtering out useless information and being overwhelmed by all the links related to that topic.

Brella - Personal Weather

Matthew Barker, Austin Astorga

Meet Brella, a beautiful and thoughtful weather app that takes the "um" out of the forecast and delivers your personalized weather forecast daily!

Winner of Ron G. Kermisch ’88 Innovation Award, 2017 Engineering Innovation Competition 

Featured by AppAdvice and Macworld!

NEW - Alexa Support

There’s a new virtual assistant in town, and one with some weather-related Personality. Just say "Alexa, ask Brella Weather..." for your personalized forecast, what the weather will be like, or just what to wear. Brella even has custom Personality phrases for her new home! Visit Brella’s settings to connect your phone to Alexa and hear just how personable some of her lines are. 

Beautiful, Thoughtful Forecast 

Brella streamlines the crowded weather forecast down to its most essential elements. Bold, vibrant colors indicate the temperature at a glance, while temperature and conditions are presented with crisp, bold text. The hourly forecast provides a clean, yet detailed look at the day’s weather, while the daily forecast makes comparing temperature ranges nearly instantaneous.

Playful Personality

While we were teaching Brella how to translate your wardrobe into a readable forecast, we might have let this super-powered AI spend a little too much time online. From pop culture and movie references to delightful messages and cheesy compliments, Brella adds that personal touch in a personalized daily forecast that will make you smile, cringe, or sminge. Look for seasonal updates!

Daily Notifications

Always checking the weather in the morning to see what to wear? Have the answer right on your lock screen when you wake up with Brella’s personalized forecast. You can choose exactly the type of information you want in your notification, from bare-bones essential weather to a detailed forecast summary. In addition, you can check out what wild wisdom Brella has to offer at any time through the Today Widget in Notification Center, or using your voice with Alexa. Note: In-app purchase required for some features.

Queue Me In: For All Your Office Hour Qs

Tiffany Wang, Karun Singh, Sophia Wang, Ryan Slama

As student enrollment in popular technical majors has grown exponentially, TA office hours have become a relatively scarce resource. Long wait times and overcrowded rooms are the new norm, and students are growing increasingly frustrated with the problem. Queue Me In is an online platform for students to join the office hour queue ahead of time and see how many others are already in line, all before having to squeeze into a tightly packed room. With this tool, students can see what others are asking about or add their own questions, and TAs can manage the crowds more effectively. For professors, we are developing a data dashboard containing analytics about the crowdedness and effectiveness of office hours, bringing real student concerns to the forefront. We hope that Queue Me In will help bridge the gap between the students in need and those who can help, making office hours a less stressful experience for all.

CiceroDB: Efficient Voice Output of Time Series Data

Mark Andrew Bryan

The interaction between user and computer is more and more shifting towards voice-based interfaces. This trend is evidenced by devices and services such as Google Home, Amazon Echo, and Apple's Siri. The Cornell Database Group is currently developing CiceroDB, an experimental database system tailored for voice-based data analysis. We demonstrate an early version of a component for interactive, voice-based analysis of large time series. A particular focus lies on the problem of data vocalization, i.e. the translation of potentially large query results into voice output. We exploit the fact that voice output has to be concise and simple to be understandable for listeners. Hence, the problem of generating voice output reduces to choosing between several coarse-grained alternatives. To make that choice, it is sufficient to evaluate the time series at a few carefully chosen locations. We use techniques from the area of optimal experimental design to choose optimal batches of sampling points. The resulting algorithm generates near-optimal voice descriptions at interactive speeds. It reads a minimal amount of data compared to exhaustive processing and generates more precise descriptions than naive sampling approaches.

Operation Bitwise

Annie Grace Hughey, Sophia Deng, Brandon Walker, Tyler Thompson, Alejandro De Vore-Oviedo

You find yourself in a strange, poorly-lit room. Complete silence, totally alone. You move towards the computer screen in front of you as it flickers. You don’t remember much about yourself, but you do know that you have deft control of computers. Within a few minutes you find yourself typing away, and before you know it, you’ve unlocked something. You’ve designed a program that translates elaborate security systems to games, allowing you to hack computers and retrieve sensitive data through solving puzzles. With the help of Vi and Russ, the viruses aiding your program, any previously impregnable system is rendered vulnerable. Take the controls, and you can beat any security system. Operation Bitwise is a slide-until-blocked puzzle game. When certain blocks collide, momentum is transferred between them in a perfectly elastic collision. The primary objective is to solve each puzzle, designed by the adversarial security system, by moving Vi and Russ to their respective goals. To beat all of the puzzles players must use Vi and Russ in concert to discover creative solutions and maneuver through dynamic challenges. As players “hack” through several computers, they are rewarded with an additional piece of the story, as well as more gameplay.

Laser Penguins

Michael Henning, Joel Hoover, Nicole Pirringer, Max Rademacher

Laser Penguins is a fluffy frenzy of photonic warfare. Battle your friends in this top-down mobile game that features intuitive, marble-style movement controls and an innovative mirror mechanic. Available for iOS and Android.

SkinnerDB: Regret-Bounded Query Evaluation via Reinforcement Learning

Deepak Maram, Samuel Moseley

Database systems use query optimization to map declarative queries to optimal execution plans. Robust query optimization becomes however illusory in the presence of correlated predicates or user-defined functions. Occasionally, the query optimizer will choose join orders whose execution time is by many orders of magnitude higher than necessary. The Cornell Database Group is developing SkinnerDB, a novel database management system that is designed from the ground up for reliable optimization and robust performance. SkinnerDB implements several adaptive query processing strategies based on reinforcement learning. We divide the execution of a query into small time periods in which different join orders are executed. Thereby, we converge to optimal join orders with regret bounds, meaning that the expected difference between actual execution time and time for an optimal join order is bounded. We demonstrate an early prototype of SkinnerDB and show that it achieves comparable performance to classical database systems on standard benchmarks. On difficult-to-optimize queries, it outperforms them by orders of magnitude.

Verifying Text Summaries of Data Sets

Niveditha Shenoy Badiadka, Naijia Fan

The Cornell Database Group is developing a novel natural language query interface, the FactChecker, aimed at text summaries of data sets. The tool focuses on natural language claims that translate into an SQL query and a claimed query result. Similar in spirit to a spell checker, the FactChecker marks up text passages that seem to be inconsistent with the actual data. This requires translating natural language claims into queries. At the heart of the system is a probabilistic model that reasons about the input document in a holistic fashion. It maps each text claim to a probability distribution over associated query translations. By efficiently executing tens to hundreds of thousands of candidate translations for a typical input document, the system can map text claims to correctness probabilities. Verification is an interactive process in which users are shown tentative results, enabling them to take corrective actions if necessary. In this demonstration, we will show how to apply the system to verify articles from various sources such as 538, NYT, or Wikipedia.

Project Samwise

Matthew Epstein, Jill Wu

Attending University should not require getting an a degree in mastering a handful of applications in order to achieve their intended degree. Samwise takes the abstract idea of an assignment, a task, or an event and makes it tangible. Samwise enables students to focus on the task at hand, while keeping them prepared for their future tasks. Project Samwise is the central hub of a student’s academic life. Managing a student’s tasks have never been easier; by being able to input a task while also knowing the assignments and tasks that he or she is currently undertaking, the student gains better insight into their current and future workloads. One of the hardest aspects of planning assignments is determining what information is important. Samwise provides pertinent fields that allows a student to organize their thoughts and priorities amid a new and sometimes confusing collegiate experience. Samwise takes the stress out of planning so students can focus on what really matters: learning.

Cornell Autonomous Bicycle

Dylan Eugene Meehan, Olivia Xiang

CU Autonomous Bicycle designs, builds, and tests a self-balancing, self-navigating, robotic bicycle to be "better" than any other. We are solving problems involving balance, navigation, localization, obstacle avoidance, and autonomous starting/stopping. We also work on a human-rideable bicycle with balance assist features.

Health Tracking App for IIH

Albert Chung, Sonia Appasamy, Ben Edwards, Eashan Garg, Katy Voor

This is a year long project to aid patients who are suffering from idiopathic intracranial hypertension (IIH) to effectively communicate health parameters to their doctors. We are in communication with a doctor specialized in IIH treatment. The project is a mobile IOS application built on React Native featuring medication reminder, headache logging, symptoms tracking, and more.

Orientation Week

David Chum, Julia Kruk

An iOS and Android app for incoming freshmen to use during their first week as a Cornell student, used by more than 2000 students during Fall 2017. The app reminds freshmen of their upcoming events, automatically selecting the required ones, and integrates with Google Maps to direct new students to the event locations. The creation of the app took a year and a team of 11, performing user research, iterating through designs, and smashing bugs, and finally marketing to the incoming freshmen. We are a subteam within Cornell Design & Tech Initiatives. The app is available on the App Store and Play Store as O-week. For more information, visit http://cornelldti.org/

CampusTour

Serge-Olivier Amega

A Cornell AppDev iOS app allowing users to explore the campus with augmented reality. Users can view upcoming events and nearby buildings.

Spyderbot

Wen-Ying Lee, Hamish Tennent, Ilan Mandel

One of the biggest design challenges in most Human Robot Interaction (HRI) research labs is finding an appropriate robot platform for the researcher to use in their studies. Most labs are made up of technically minded engineers and human focused social scientists. When looking to research the effects of new behaviours/algorithims/systems often the only choice is to use one of the expensive, off the shelf products available on the market (eg. Keepon, Nao, Pepper, etc). This means new interactions are being limited by what can be designed into these existing form factors using limited access to the hardware platforms themselves. This has the effects of limiting the generalisability of scientific research and I believe is a choke point for the robotics innovation pipeline. When so many of the findings published today are either using heavily anthropomorphised products already available on the market (e.g Nao robot), products with limited access the the backend software (e.g Jibo), or low fidelity robots prototyped over a weekend in a research lab, it creates a gap in the explorable knowledge for the field of HRI. The goal for this project was to design a non-anthropomorphic, open source, low cost, expressive social robot platform that could be widely used. My goal was not only have it used by the 6+ researchers in my lab, but those in other departments, institutions and with the goal of doing field research with end users to explore potential use cases for social robots in the home. This work is still ongoing and this robot is currently being used in research in the lab. Usability tests are being run to verify whether or not the design intentions were realised in the final design. The robot is also being used in research into how people learn a new language, this is allowing practical testing of the usability of the form to convey expressive behaviours.

Cornell Design & Tech Initiative

Karun Singh, Alice Pham, Matt Barker, Shea Belsky

We are a project team that creates technology to address needs on Cornell' campus, and beyond. We currently have over 50 members (designers, developers, and product managers) actively collaborating on 9 products that are centered around the Cornell community. Through these projects, we are looking to improve things like campus safety, orientation week, office hours, undergraduate research, and more!

Shout

Neha Rao, Raymone Radi, Kaushik Ravikumar, and Kaitlyn Son

Shout is an anonymous sexual assault reporting app that looks to give otherwise silenced survivors a voice in the Cornell community. The reports are used to warn other students of dangerous areas on and around campus.