COZIE
Sleep deficiency is a growing concern in modern society. Studies have shown that alternating human skin temperature between certain values actually increases the quality of sleep for a user. Cozie automatically regulates a user’s body temperature before and during sleep to enhance the duration of uninterrupted sleep. A user’s temperature and movement are captured by sensors embedded within a heated blanket, and are used to determine the user’s sleep state. The data is then processed by a controller, which actuates the heating system.
Group members: Christopher Stern, Charlotte Chan, Vishal Mathur, Nathan Yim
DRIVEN
This project aims to detect the driving behaviours of checking blind spots and making lane changes using an inward facing camera and gyroscope. The data from the sensors are collected and processed offline. Machine learning techniques are used to identify the start and end times of each driving behaviour as well as classify each behaviour as being either good or bad.
Group members: Christopher Ngan, Joshua Cheung, Angela Gu, Clement Fung
ALOHOMARA
Metal keys have been around for years. They should be things of the past, but you still see people scrambling to look for their keys in front of their apartment doors. Electronic combination locks are being used in some places, but they still don’t solve the problem of difficulties in key sharing. Working with mobile application, Arduino, custom transreceiver and other technologies, Team Alohomora built a smart lock that makes unlocking doors and sharing keys with others easier than ever.
Group members: Clayton Hung, Rochelle Chong, Andy Seo, Bix Deng
MEDICAL IMAGE SEARCH (MIS)
Accurate image retrieval for the medical industry has yet to be realized. Traditional image search methods use meta and header data from images to retrieve the correct result. Medical images only have high level and broad data that would not yield accurate results. A content based image search method is presented that uses radon transforms to index and search for images, and has outperformed the state-of-the-art in speed and accuracy. In the future, it may be integrated into a hospital’s network for “virtual peer review”.
Group members: Sumanyu Sharma, Laura Ospina, Davin Wong, Imaad Umar
LIFTSMART
LiftSmart is a workout-assisting device that helps beginner and intermediate level weightlifters improve their form and technique for the conventional squat and deadlift. The system is composed of Inertial Measurement Unit (IMU) sensors attached to the user’s back and thigh, connected wirelessly with a smartphone app that the user can monitor their performance with. Based on the user’s progress, the app will provide both short-term and long-term suggestions and methods for improving lifting form and technique.
Group members: Tianlei Hou, Rakesh Rajkumar, Buvan Srinathan, Alex Yau
MOZART
The Mozart team has been working on improving the user interface for a stroke rehabilitation robot for Professor McPhee and PhD candidate Borna Ghannadi. The Quanser stroke rehabilitation robot assists stroke patients to regain mobility in their upper limbs. The team’s goals are to improve the physical interface of the robot and to create an interactive software experience for the user. The end goal of the project is to have an ergonomically designed interface for rehabilitating stroke patients with a gamification aspect to optimize the experience.
Group members: Alan Nelson, Zack Bernholtz, Sofia Loureiro & Mozart, Hojun Pyo
MASSMEDIC
Mass casualty incidents (MCIs), like shootings and earthquakes, are particularly difficult for emergency services to respond to due to resource and information constraints. One emerging source of valuable information for MCIs are Tweets and other real-time crowdsourced posts by victims and bystanders. MassMedic takes this one step further, facilitating automated collection of critical victim data through smartphones and wearables, allowing paramedics to visualize and act on triage information from MCIs earlier and more efficiently than possible today.
Group members: Mayank Gulati, Lesia Nalepa, Nick Mostowich, Harvey Tang
TOP SECRET
Inefficiencies in the grocery checkout process create long wait times which are detrimental for store’s returns Laser scanning requires a specific orientation and closeness of an object limiting the speed for manually scanning. This projects looks to become a semi-automated system that replaces lasers with an image processing solution. Multiple cameras capture all sides of any object passing through the system, thereby ensuring detection of its barcode. By using well-thought usability and efficiency benchmarks, the solution seeks to replace both cashiers and the self-checkout systems.
Group members: Noorain Noorani, Jeffrey Rodrigues, Yannick Ngana, Arthur Yan
GRAND RIVER TRANSIT PERFORMANCE DASHBOARD (EAGLE)
EAGLE is a web-based transit planning and operations support tool developed for Grand River Transit (GRT) used to address complaint-driven transit problems, outdated data presentations, and limited navigation capabilities of GRT’s current platform. Leveraging Automatic Passenger Counting (APC) systems, EAGLE focuses on clearly organizing and displaying existing issues in bus routes pertaining to ridership activity and operational capacities, with an array of useful filtering capabilities to help transit planners proactively see potential issues in the route network.
Group members: Andrew Lee, Tracy Lo, Solji Han, Ethan Li
KEY VALUE STORE
Key Value Store is a key storage solution that improves the process of key transfer between property owners and renters. A property owner will place their key within a Key Value Store system. The system will then lock leaving the key accessible by passcode. The renter can pick up the key at their convenience by entering the correct passcode. The key’s profile is recognized by the system, and so the owner will be notified, via a mobile application, if the renter returns a faulty key.
Group members: Kai Wang, Shuai Huang, Terence Kim, Ryan Ong, Eugene Jeong
HOCUS FOCUS – DYNAMIC BIFOCALS
Multifocal lenses such as bifocals and progressive lenses provide an inadequate user experience due to limited field of vision, lens distortion, and ergonomic issues. The team presents a dynamic variable focus system that improves upon the traditional lenses by automating a pair of Alvarez lenses to mimic human eye accommodation. The system computes the fixation distance from pupil movement and adjusts lenses to the user’s distance or reading prescription as needed.
Group members: Raphael Kalker, Anni He, Timothy Jeruzalski, Ryan Leslie
AUTO MIT
Aphasia is a collection of language disorders caused by damage to the brain, resulting in the inability to comprehend or reproduce language, or both, with various levels of severity. Current day treatment methods currently have a variety of setbacks that deter treatment: cost, availability, social stigma, etc. The purpose of this project is to create an alternative treatment method for a subset of Aphasia, by developing an automated system of deploying Melodic Intonation Therapy, to treat language reproduction issues.
Group members: Jerome Couture, Ryan Nhieu, Thomas Kelly, Allan Cao
PEACHES
Many fruits are delicate and must be harvested manually; however the farming industry often lacks manpower to harvest the fruit, causing food waste. We are creating an autonomous robot to help alleviate the manpower needs; by designing a soft robotic peach picker, we are designing for one of the most delicate fruits, making it easily adaptable for other types of soft fruits. Our project incorporates aspects of image processing and mechanical engineering along with a soft robotic gripper to achieve this goal.
Group members: Ian Easterbrook, Daniel Saunders, Matthew van Gennip, Prarthana Chandran
FALCN
Group members: Jordan Wyatt, Jacinta Ferrant, Victor Vucicevich, Adam Wootton
MUSE
Muse is aimed at helping museum administrators and curators collect analytics on visitor-exhibit interaction. The system tracks visitors through their visit using a mobile web app and records the popularity and holding power of exhibit elements as well as the path taken by each visitor through an exhibit. The app also provides the user with additional information about each element. The resulting analytics are made available via website dashboard to help administrators and curators gather relevant data for designing future exhibits or to justify funding.
Group members: Taylor Martin, Lindsay Paterson, Andrew Illmann, Evelyn Zhang
TRACKCONDUCTR
Due to financial limitations and/or space constraints, many theatres are unable to hire enough musicians to cover every part in a musical’s orchestration. TrackConductr aims to solve this problem by allowing conductors to control the playback of pre-recorded audio tracks, complementing live musicians and giving a fuller sound. Using image processing, TrackConductr extracts the tempo from a conductor’s natural gestures in real-time and dynamically plays a MIDI file.
Group members: Gurneet Arora, Nicole Simone, Melissa Lynett, Canmanie Ponnambalam
BCI WHEELCHAIR
This design team is developing a training simulation for severely disabled users (e.g. those with ALS, paralysis, tetraplegia, etc.) to control a wheelchair using mental commands. This brain-computer interface (BCI) takes the user’s thought as input for the wheelchair’s movement. The simulation allows the user to control their movement in a virtual environment and test the system’s streamlined control scheme and the obstacle avoidance measures.
Group members: Theo Chan, Danielle Johnston, Chelsea Haemel, Jim Lin
WALK-E
WALK-E is an autonomous rollator retrieval system. It is designed for the independent senior in mind. When the rollator has been placed out of the physical reach of the user, the WALK-E system operates to autonomously locate the user and drive it to them. The system is composed of an adjustable mechanical aluminum frame. Attached to the frame are motors which provide actuation of the drive wheel. Also are attached a Kinect sensor, Inertial Measurement Unit (IMU) and a laptop running Robot Operating System (ROS).
Group members: Hazik Afzal (SYDE), Kishan Dedakia (SYDE), Ben Vander Schaaf (MTE), Jackson Bush (ME)
WALL_E: THE GREEN ENERGY WALL
With an increasing demand for clean energy there is a great need for renewable energy enabling technology. Used electric vehicle batteries creates an opportunity for use in Energy Storage Systems (ESS). Our FYDP demonstrates how an ESS can provide cost savings for home-owners/business by shifting electricity purchases to cheaper off-peak times. For utilities, our system can support renewable energy generation and regulate demand. Our project demonstrates the capacity and overall performance of re-purposed batteries in addition to the development of policy strategies for their use in Canada.
Group members: Julia Kavuma, Marwan Saadeldin, Andrew Andrade, Alen Daniel, Wajeeh Syed