Facilitate Personal and Community Connection

2020-2021 Capstone Design Projects

Project Name: USB radio for networking remote health centres

Project Client: Signalytic

Technology: Computer & Electronics

Project Description: Our client, Signalytic, aims to create health data information systems in Uganda by having Internet connected microservers placed at healthcare centers to form a network that reflects real-time and correct medical data. But rural healthcare centers cannot connect to this network due to non-existent Internet access and limited power. So the microservers in the rural healthcare centers need to have additional hardware to gain access to the information shared on the network.

Our solution is to use USB connected transceiver dongles that provide a long range wireless link between a rural microserver and Internet connected microserver. Our design extends the range of Internet access through the dongle’s antenna of an Internet connected microserver and is subsequently picked up by the dongle’s antenna from the rural microserver, allowing data exchange between the two servers. Our solution will allow a wireless connection to be established between rural servers up to 5 kilometers away, allowing digital paper based files to be exchanged to the real time network at a relatively fast rate. Ultimately, rural healthcare centers will be connected to the information system that provides real-time medical data for health care workers to save those in need.


Project Name: Development of Arm Cycle Control to use with XBox Racing Games by Disability Communities

Project Client: Physical Activity Research Centre (PARC) at ICORD

Technology: Computer & Electronics

Student Contacts: Scott Beaulieu [nicholassjbeaulieu@protonmail.com], Edward Luo [luohaoubc@gmail.com], Keith Consolacion [keithcon@student.ubc.ca], Fabian Lozano [fabian.lozano@protonmail.com]

Project Description: Individuals with physical disabilities often have limited options to be physically active. A popular form of cardio exercise for people with Spinal Cord Injuries (SCIs) is ‘hand cycling’. At the Physical Activity Research Centre (PARC), users with SCIs can engage in many physical exercise activities while building social connections within their community.

With the ongoing COVID-19 situation and the initial closure of PARC, it was clear that there was a need for a new home-based exercise device for individuals with SCIs. The goal of this Capstone project was to develop an Arm Cycle Controller, compatible with the Xbox adaptive controller for Xbox One Series consoles and Windows 10 PCs. The final product is aimed at bringing an inexpensive and engaging exercise experience into the end user’s home, one that can foster community through online engagement.

Our solution is an electromechanical adapter that adds controller functionality to a commercial mini-exercise bicycle. The device achieves this by sending control signals derived from encoder and inertial measurements to an Xbox Adaptive Controller through audio cables. These signals are interpreted as joystick and trigger inputs which allows for racing games to be played using an Xbox One Series console or Windows 10 PC.


Project Name: ED2Foundry Application

Project Client: St Paul’s Hospital and UBC Emergency Medicine

Technology: Digital Media, Web & Mobile Apps

Project Description: Foundry is a province-wide network of integrated health and social service centres for young people aged 12-24. It provides a one-stop-shop for young people to access mental health care, substance use services, primary care, social services, and youth and family peer support.

Physicians across BC refer patients to Foundry on a daily basis. The current referral system has two major problems.

Firstly, they utilize an obsolete paper pamphlet system. These pamphlets are easily lost or misplaced. Moreover, the information on the pamphlets is not personalized to the individual youth patient. This makes it difficult for them to locate and correctly utilize the information provided.

Secondly, the current system provides no means for physicians and Foundry clinicians to exchange information regarding individual patient cases. This connection between physicians and clinicians can greatly improve the quality of service provided to the youth.

We developed a mobile application to resolve the issues. This app allows physicians to refer youth patients via SMS or email, a much more convenient communication channel with the ability to personalize information. Physicians and Foundry clinicians will be able to share patient progress on the app to improve the provided quality of service.


Project Name: Game Narrative Development ToolKit

Project Client: Blackbird Interactive

Technology: Software

Project Description: The video game industry is currently experiencing rapid growth. In 2020, the video game industry grew 12% to a total of 139.9 billion dollars. As video games become increasingly popular, there will be more competition to develop more complex and immersive games. Dialogue lines form a core experience for any great video game and more complex games can have tens or hundreds of thousands of lines. Traditionally, video game development studios have had to manually track these lines, however this becomes increasingly difficult as games grow larger. The Game Narrative Toolkit provides a simple way for video game development studios to track dialogue lines for their games over the course of development, from writing to revisions to recording. Users have access to an intuitive application that allows them to import screenplays, export lines into a voice over script, and view hundreds of thousands of lines, including any revisions made during the development process. With the Game Narrative Toolkit, game development studios no longer have to rely on outdated processes to track dialogue lines. Instead, they can focus on making the games that they want to make.


Project Name: Immersio Language Learning Mobile Application

Project Client: Immersio Learning Incorporated

Technology: Software

Project Description: Wisdom stemming from ancient and indigenous languages is being overlooked due to a lack of opportunities for conversational language learning. Though interactive learning platforms currently exist for modern languages such as Spanish and French, there is a lack of options for ancient languages. Immersio fills that need with their web app.

These days most people use their phones on a daily basis. Immersio was challenged with having no way to access these users who would rather use a mobile app than a web app. So we developed a mobile app where users can learn Latin on the go at a difficulty personalized to their needs and goals.

The key feature of the project is the offering of comprehensive Latin lessons of varying complexity, which can be set by the user. Supplemental practice exercises that utilize both audio and visual cues are also included to augment the learning experience. A particularly challenging aspect of the project was the integration with a custom built natural language processing (NLP) model in the form of a Rasa X chat bot to facilitate conversational learning. Lastly, we also incorporated Google Authentication as part of our user progress tracking infrastructure, which consists primarily of our users and lessons databases hosted in Microsoft Azure.


Project Name: Mitigating social isolation through integrated Interest and geo-location matching software

Project Client: Proximy Technologies Inc.

Technology: Digital Media, Web & Mobile Apps

Project Description: Our project addresses social isolation, a global issue that greatly affects young adults. Social isolation has been amplified by the pandemic and is associated with increased physical and mental health risks. Proximy, a local, Vancouver based company, aims to end social isolation while promoting individuality through its web and mobile applications. Our main deliverable is to design an intuitive user experience and implement an Android mobile application that supports discovering nearby users with similar interests, interesting events, and chatting with the community.

Our main design challenge is to design a prototype mobile application that is scalable, accurate, fast and maintainable. This was done by considering various options for the backend, such as Firebase, and front end, such as Flutter, and designing clean and efficient software. The overall system and geo-matching algorithm is also designed to scale with a larger user base while also maintaining real time updates for notification and messaging features. Another challenge was ensuring that common user journeys were intuitive, enticing, and error-free, and integrated well with the web application.


Project Name: Smart Advertising Campaign Generator

Project Client: Adsight

Technology: Digital Media, Web & Mobile Apps

Student Contacts: Victor Xiong – https://www.linkedin.com/in/victor-xiong-791137149/, Eddy Maric – https://www.linkedin.com/in/eddy-maric-233b67174/, Michael Milic – https://www.linkedin.com/in/michaelmilic/, Shuai Lan – https://www.linkedin.com/in/shuai-lan-457962207/, Rex Dong – https://www.linkedin.com/in/rexdong/

Project Description: The Smart Advertising Campaign Generator is a project for Adsight – a start-up company disrupting the outdoor advertising industry. Two primary industry problems are: 1) it is difficult for ad buyers to calculate the return on outdoor advertising; 2) outdoor advertising properties are owned by a variety of media supplier companies which can complicate the booking process to rent ad space.

The Campaign Generator addresses these problems for prospective ad buyers (small businesses/marketers) by 1) providing them actionable insights on advertising properties; 2) consolidating the ad space booking process in one platform that recommends advertising properties based on ad buyer personalized needs.

Adsight launched a minimal viable product – a front end React application powered by Firebase. We were tasked with migrating it away from Firebase by creating an API. We accomplished this by making a NestJS backend, dockerizing the application, and hosting it with AWS.

The Campaign Generator also posed a technical challenge. We wanted to treat inventory recommendations as a multi-label machine learning classification problem, but didn’t have any data to train a model. Instead, we created a constraint optimization algorithm to recommend inventories and saved the results for future machine learning endeavors pursued by Adsight.


Project Name: The Artist’s Auction House

Project Client: Art Vancouver Productions Inc.

Technology: Software

Project Description: Art Vancouver is Western Canada’s largest contemporary art fair which aims to bring together both growing talents and renowned international artists under one roof.
However, due to the restrictions caused by the COVID-19 pandemic, the 2020 Art Vancouver fair had to be postponed until further notice.

In order to find an optimal solution, Art Vancouver partnered up with UBC Capstone Team 47 to develop an online platform where artists around the world can attend the art fair virtually.

This project, “The Artist’s Auction House” is a website created to unite both artists and art buyers.

The website applies simple, yet visually appealing design in-line with the Art Vancouver aesthetic to capture the attention of its users. With a robust backend service, users of the website can browse through thousands of artworks from anywhere at any time. After a quick sign-up, users can create listings, buy artworks, or bid on auctions with just a few clicks. The purchasing process is also streamlined so users have guidance every step of the way.
The Artist’s Auction House is our way of providing support in these difficult times so artists can connect with each other, inspire one another, and educate aspiring young talents.


Project Name: XR iNTERACT Multiplayer Sports Game

Project Client: Texavie Technologies Inc

Technology: Computer & Electronics

Student Contacts: Robert Migut: robmig626@gmail.com, Ryan Homsani: homsani@protonmail.com, Yuyi Wang: yuyiw1121@gmail.com, Brennan Cathcart: brennancathcart@gmail.com

Project Description: We worked with Texavie, a Vancouver-based tech startup specializing in developing smart-apparel and wearable devices. Their mission is to become the leading company for movement data, health analytics, and immersive virtual user experiences. Our capstone project specifically targets Texavie’s short term goal of gaining traction in the video game industry. Before our project, Texavie did not have an interactive way to demonstrate their technologies’ ability to provide an immersive video game experience to stakeholders, potential investors, and video game players. We resolved this by incorporating Texavie’s
iNTERACTTM motion capture hardware into a fun, engaging dodgeball experience. Players control the video game using the same actions as if they were playing dodgeball in real life such as picking up a ball, throwing a ball, and ducking to dodge and incoming ball. Texavie’s hardware is still in development so we hope to see continual evolvement of this project beyond the scope of the capstone course.


Project Name: Automated Traffic Police Signal Recognition System

Project Client: Huawei Technologies Canada Co. Ltd

Technology: Software

Student Contacts: ece491group30@gmail.com

Project Description: Currently, autonomous vehicles analyze road signs and obstacles to safely navigate. However, there are no means for a vehicle to understand a traffic controller at an intersection due to the dynamic nature of the interaction. The Automated Traffic Police Signal Recognition System project aims to bridge the gap in human-machine interaction for self-driving cars.

The project leverages Huawei’s Ascend hardware platform to do head pose estimation body pose recognition in real time. The prototype utilizes the head pose estimation model to make sure if the traffic controller’s head is facing the car so it can start listening for directions. Upon verifying the directions are intended for it, the body pose model estimates the gesture directed by the traffic controller which will drive the motors of the RC car. The prototype is capable of understanding directions for turning left, turning right, stopping, and driving straight. The machine learning prediction results are streamed in real time to a presenter server using Wi-Fi.

The team hopes to further human-machine interaction research in the autonomous vehicle industry with our open-source contribution. Thanks to Huawei and UBC ECE for sponsoring this capstone project. The team can be contacted at ece491group30@gmail.com


Project Name: Hand-Gesture Controlled Robot Pet

Project Client: Huawei Technologies Canada Co. Ltd

Technology: Software

Project Description: Consider having a pet that is smart, responds to your commands without training, and even takes pictures for you! You don’t even have to pick up after or buy kibble for this pet! This is a dream come true for those craving an animal companion but cannot afford the time, money, or maintenance required for real animals. Robot pets are one example of how technology can assist in scenarios where companionship and fun are needed, but the handling and upkeep of real pets isn’t possible; especially in healthcare or senior homes.

Apart from doing normal pet things such as moving around and performing a few tricks, it can also perform robot tasks, like capturing photos with the camera in its nose. It can also communicate its mood through animated eyes on a colored LCD screen mounted to its head.

Through its camera “eyes”, this robot pet uses a combination of machine learning and computer vision to see and respond to commands given through its owner’s hand gestures. The robot pet can track, recognize, and perform a task such as following the user while video capturing. The user is thus able to control the pet from a distance, enabling full autonomy.


Project Name: Human Anatomy Teaching APP

Project Client: UBC Faculty of Medicine

Technology: Digital Media, Web & Mobile Apps

Project Description: The Human Anatomy Teaching App is a mobile application which provides students with an out of classroom experience to learn human anatomy with real cadaver images. This application makes use of interactive quizzes, flashcards, video references, and labeled inspectable cadaver images. This is especially important during times like this where access to in person labs are not available. This also provides students with additional learning material they are able to access from anywhere.

The goal in this iteration of the application was to improve the overall student learning experience, and expand upon teaching capabilities of the app. The major changes that have been made to meet these goals are the implementation of synchronized user progress tracking, fill in the blanks style questions, search and filter functionalities, and enhanced zooming. Progress tracking with positive reinforcement helps students to get real time feedback on their learning progression, and helps to motivate them to continue improving. Fill in the blanks questions allow for recollection learning rather than just recognition learning. Searching and filtering content in the application helps students find exactly what they are looking for when studying. The enhanced zooming gives students a more accessible and clear zooming functionality that is vital when closely inspecting cadaver images.