Enhance How We Do Things

2020-2021 Capstone Design Projects

Project Name: Integrate RFID into seafood manufacturing software

Project Client: ThisFish

Technology: Computer & Electronics

Project Description: Our project is mainly devoted to help our client (ThisFish) with tracking units of their product as it goes through all their stages in production. Through our efforts, we not only implemented the RFID scanning system for the blind spots, but also added new functions such as database, data visualization and data prediction to facilitate the employee efficiency statistics of our customers. During the design process, we solved the networking communication problem between the RFID system and client’s software backend by letting the RFID hardware talks to our system first on a local server and then use the server to communicate with Thisfish’s backend. In subsequent database development, we address the issue of on-time data updates and data presentation with a timing function and a shifting logic. Finally, after the completion of our design, customers can use our system to achieve daily, weekly, monthly and annual statistics of employee data, visualization of employee data and prediction of employee work efficiency, which will help our customers to better manage employees.

Project Name: “Drive-It” Bike Assist – Compact Hybrid E-bike System with highly efficient regen/battery power boost and integrated wireless Smartphone app

Project Client: Geyer Engineering Ltd.

Technology: Digital Media, Web & Mobile Apps

Student Contacts: Anav Chopra (anavch@student.ubc.ca), John Shen (johnmone@student.ubc.ca), Pablo Fernandez (pablofer@student.ubc.ca) Takavada Chigumete (chigumet@student.ubc.ca), William Luo (dehui15@student.ubc.ca)

Project Description: As e-bikes experience a surge in popularity their limitations do not. For instance, they are bulky, owing to their integrated nature. They come with a massive motor attached to either the front or the back wheel and a battery pack that is attached to the main body of the bike. Our team decided to tackle these limitations by creating an ‘e-bike conversion kit’ which allows the user to convert their normal bicycle into an e-bike and go back to being a regular bicycle as and when they please. We also included regenerative braking in our conversion kit so that the battery can be kept very small and compact, taking away the heavy and tough to move nature of a standard e-bike. Using our custom-built assembly that can be attached to the rear wheel of the bike, reduces the weight contribution of the motor on the overall frame of the bike and allows for a smooth riding experience.

Project Name: Augmented Indoor Position Tracking App

Project Client: Delta Controls Inc.

Technology: Digital Media, Web & Mobile Apps

Student Contacts: Artur Danilenko, Harrison Wu, Joel Ritter, Samin Islam, Tyler Mann

Project Description: The purpose of the Augmented Position Tracking App project was to create a mobile application that interfaced with a network of sensors through out a building to display the current position of the user within the building on a google floor plan UI, and also for the user to be able to see information about the environmental state of the areas around them from the map. The app is designed to work with the O3 Sensor Hub 2.0 made by Delta Controls. The app aims to help people to navigate large buildings by showing their real time position on the map, and to help people who are looking for particular spaces, for example to study in, have a meeting in, etc., by allowing the user to filter areas based on designed range of temperature, humidity, brightness and noise. The major technical hurdles that the team had to overcome were deriving the user’s location based on the signal strength of the Bluetooth beacons of nearby sensor hubs, communicating with the O3 sensor hubs to receive their sensor readings data and building a google maps integrated mobile application.

Project Name: City-Wide Multi Reusable Container Platform

Project Client: Natural Source/ShareWares

Technology: Software

Project Description: In 2018, of the 64% material that was diverted from disposal in Metro Vancouver, merely 4% was able to be reused for its original purposes. The rest? Sent to landfills.

ShareWares is a container-sharing company based in Vancouver that plans to implement a closed-loop container sharing program in Metro Vancouver. The mission of ShareWares is to promote zero waste in the community by creating a scalable and duplicatable platform for a city wide reusable ware system. This circular economy will allow for the sharing of a variety of ware types across industries such as cups, food containers, and meal-kit packaging. To achieve this goal, ShareWares will play an important role by connecting the retailers, the sanitation warehouse, the container manufacturer, and the end users through their closed-loop system.

The role of our UBC Capstone team was to implement technology for ShareWares that supports their closed-loop program. This includes a technology-enabled bin, technology-enabled scanning apparatus, database, and web application. Using ID tracking technology, cloud server database, scanning technology, all these components work together to make up a system to support SharesWares closed-loop reusable container program.

Project Name: 360° Traffic Sensor

Project Client: Breeze Labs Inc.

Technology: Computer & Electronics

Student Contacts: CapstoneTeam87UBC@gmail.com

Project Description: Our client Breeze Labs aims to build a solution that helps cities reduce CO2 emissions by using AI framework to measure and optimize traffic flows at intersections.

However, they encountered difficulties using the existing camera technologies as they need to retrain their AI model for every type of sensor it encounters. This approach is both inefficient and time consuming. Breeze Labs needs a universal device that is versatile enough to detect different types of vehicles (cars, buses, trucks, bicycles, etc.) while retaining compatibility with its AI framework.

We delivered a versatile, robust, and affordable camera prototype for object perception and tracking at intersections. Vertically adjustable cameras provide a wide coverage angle and are easily modifiable through a web user interface (UI). The four-camera-one-microcontroller design encodes and streams video data in real time, allowing our client’s AI framework to operate optimally. Moreover, the stream and the UI are regulated automatically by an embedded system to ensure system reliability.

For more details, please check out our promotional video: https://www.youtube.com/watch?v=pwSXI65Y2Co&ab_channel=Team87CapstoneUBC

Project Name: Open Loop Directional Comparison Blocking (DCB) Scheme for BC Hydro’s Overhead Distribution System Designated as the Gulf

Project Client: BC Hydro

Technology: High Voltage Electrical

Student Contacts: Brysen Watson – watson.bm4@gmail.com, Jason Ong – jasonong.175@gmail.com, Jung Choi – wjdfkr22@gmail.com, Christopher Wan – chr.xiaow890@gmail.com, Kareem Moawad – kabdelraouf@gmail.com

Project Description: Our team was assembled to research and design a Directional Comparison Blocking (DCB) scheme for BC Hydro’s Overhead Distribution System designated as the Gulf Island Loop. The current electrical grid was set up years ago with a peer to peer loop control system between the automated reclosers of the loop and is nearing its operational life-span limit. The successful implementation of a more advanced DCB-based protection scheme for the Gulf Island’s overhead open loop would replace the existing protection scheme. This would increase the grid’s reliability as the gulf island experiences frequent power outages due to weather conditions.

We were provided with a list of desired solutions for each of the telecommunications and relay selections to achieve our goal. Utilizing Structured Decision Making (SDM) tables we narrowed down our selection to one relay and corresponding communication platform. Making use of combinational logic, we created a DCB scheme for the protection that requires minimal human input when changing the open point of the loop.

Project Name: UBC Track and Field Analytics

Project Client: UBC Track and Field

Technology: Software

Project Description: Athytics is a software tool that will be used by the UBC track and field team to help guide their scouting decisions. The vision for this project was to write software that can be built upon on the future. Before Athlytics, scouts would manually search across multiple websites for athlete placement data. Athlytics hosts, organizes and even visualizes these data points. A core feature requested by our client was the ability to predict how well a high-school athlete could perform in university using data from past and present athletes. Predicting how well an athlete could perform was the most technically challenging aspect of the project. The team spent time testing various algorithms including machine learning algorithms similar to the Netflix recommendation system. The team found there was not enough data to use a machine learning model so we decided solve this problem using the Inverse Mean Square Error.

Project Name: Real-time training load monitoring in sports

Project Client: Canadian Sport Institute Pacific

Technology: Canadian Sport Institute Pacific

Student Contacts: Harry Byrne – harrybyrne19@gmail.com

Project Description: Currently, high-performance sporting organizations rely on slow and inefficient methods to determine athlete load. This lack of real-time data for coaches can lead to athlete overtraining and injury. Our product aims to streamline this process and provide real-time load data to coaches and their athletes.

To accomplish this, our team worked alongside the Canadian Sports Institute Pacific to create a system consisting of inertial measurement units (IMU) and an offline iOS app. This system will gather, process, and display sports metrics for volleyball, wheelchair rugby, and snow sports in real-time.

To select the IMU for our system, our team compared and tested three pre-existing IMUs. Our iOS application was then designed around the selected IMU for the three sporting disciplines with the ability to calculate and display data on jump height, jump count, distance traveled, and time spent at predetermined speed thresholds from multiple IMU concurrently. Our application also allows coaches to record these metrics and export/download them to/from a dropbox account.

We hope our capstone project can be a part of helping high-caliber athletes and coaches achieve their performance goals.

Project Name: Capturing and Synthesizing UBC’s Tree Inventory for Biodiversity Enhancement

Project Client: SEEDS Sustainability Program/UBC Campus and Community Planning

Technology: Computer & Electronics

Project Description: The UBC Vancouver Campus is home to an urban forest of over 18000 trees and more than 170 different species. As the need for sustainable urban development increases, it’s becoming more important than ever to systematically analyze UBC’s trees and their ecological benefits. Capstone team 117, in collaboration with UBC Campus and Community Planning and UBC SEEDS, has developed the Tree Ecosystem Services Analysis Platform (TREESAP), a tool for analysing and visualizing ecosystem services on the UBC Vancouver campus. We have built a processing pipeline that extracts locations of tree cover on campus from aerial scans. We also provide an app to visualize this data, allowing users to generate reports about the carbon sequestration and avoided rainwater runoff that trees provide. This app can be used to plan future developments by modeling the effects of adding and removing trees, as well as to visualize the difference in tree cover between years. TREESAP aims to provide baseline data for ecosystem services on campus, helping urban planners develop quantitative arboreal sustainability goals that support the sustainable development policies at UBC.

Project Name: Design 3D Map and Flight Missions for a Passenger Drone

Project Client: Westdrone Inc.

Technology: Software

Student Contacts: Alexander Betts (xanderbetts@gmail.com), Leon Wang (leonwang1112@gmail.com), Andy Zhou (haonanzhou1998@gmail.com)

Project Description: Westdrone Flight Control is an Android application that runs on a tablet computer. It interfaces with the pilot of a VTOL, vertical takeoff and landing aircraft, by providing flight navigation assistance. Flight Control provides electronic mapping and pathing which is currently not included in most VTOL navigation systems.

There are three modes of operation: Explore Map, Simulation, and Real GPS. During  real flights, the pilot selects Real GPS. This mode enables the external GPS, marking the pilot’s location on the map.The application’s map displays a flight path with waypoints, providing real time flight statistics, and warning the pilot if the VTOL is off course. Landmarks and other important markings, such as airspace rings and Canada-US border, are also displayed.

Flight missions are customizable using 4 toggle buttons in the advanced settings menu. These enable/disable certain features. The toggles include Satellite View, Boundary Outline, Landmarks, and Off Course Warnings.

The pilot can also view flight mission history to compare the actual path taken versus the recommended path. This feature helps the pilot in redesigning future flight paths based on historical data.

Project Name: Reusable Mug Tracking and Exchange Solution for a Circular Sharing Program

Project Client: mugshare

Technology: Digital Media, Web & Mobile Apps

Project Description: Mugshare is a reusable mug sharing program initiated by students at UBC aiming to reduce the single-use cup consumption. The mugshare program was piloted at cafes across UBC in early 2019. Some problems in the pilot launch of the program were the loss of mugs, the complicated mug borrow and return process and the low user loyalty to the program. ECE Capstone Team 104 designed a QR code based mug tracking system for mugshare. When users borrow and return mugs at cafes, the baristas will scan the QR code of the mug and register the transaction to the back-end system. The cafe managers and mugshare administrators can query data from the back-end system gain an insight in the to operation status. Users can also download a mugshare and show a user QR code to baristas when borrowing mugs. Users can get deposits returned to their account when returning mugs and use the deposit in the future. Also, a contact-less return bin was designed to simplify the mug returning process. Users who borrowed mugs with their mugshare user account can return mugs to the bin directly and skip the line at cafe counters.

Project Name: A novel system for multimedia discovery in Instagram based on Hashtag optimization

Project Client: BroadbandTV Corp

Technology: Software

Student Contacts: Ash Tan – ashtan19@gmail.com, James Li – lijialu1009@gmail.com, Kaichen Wang – wangkaichen970420@gmail.com, Kurtis Chan – kurtis.chan.1@live.com, Mark Pang – markpang604@gmail.com

Project Description: Hashtags play a huge role for content discovery on social media platforms like Instagram and Twitter, but the process of generating hashtags is still manual & time-consuming.

The goal of this project is to build hashtag recommendation system for Instagram aimed at generating hashtags optimised for increasing impressions of the post. The project uses an ensemble of state-of-the-art deep learning neural networks to generate hashtags that are most relevant for an Instagram image. The contributions of the team range from understanding data collection pipeline, building a system to collect training data, experimentation with various computer vision and multi-modal techniques to generate most optimised hashtags and API development of the solution.

Project Name: Atmospheric Thermal Sensing for Autonomous Unmanned Aerial Vehicle Navigation

Project Client: JHN Aerospace LLP

Technology: Software

Project Description: Every day, the Earth’s surface is heated by the sun. The air pressure, temperature, and moisture differences that arise from uneven heating are what create our weather. Specific weather conditions and environments allow more efficient flight methods, such as soaring. In fact, soaring is commonly used by birds to fly higher without having to flap their wings. This phenomena increases bird flight duration by minimizing the energy that they use. Inspired by nature, glider pilots also use soaring to increase flight duration as it lowers reliance on fuel. Currently, finding the necessary environmental conditions for soaring relies heavily on the pilot’s experience and expertise, having up-to-date knowledge of weather conditions, and using a vertical air speed sensor on the aircraft.

Finding a more efficient and reliable system for detecting soaring environmental conditions will allow remotely piloted aircraft systems to fly longer, farther, and use less energy. The TL100 Capstone Team has worked closely with their client, JHN Aerospace, to develop a technology capable of remotely detecting the necessary conditions for soaring. If successful, reliable soaring for remotely piloted aircrafts is not too far away!