Justine Dy, Tina Chen, Michelle Xiong, Jega Ganesh
Studies have shown handwriting notes improves retention and deepens understanding of a topic. The objective of this project is to develop applications that are capable of converting handwritten text into an editable digital copy while adapting to one’s handwriting style. Once digitized, the user can easily manage and access their notes electronically. This project leverages knowledge of distributed systems, databases, optical character recognition algorithms, and cross-platform development. The main advantage of this solution is being able to capture and edit handwritten text efficiently.
SMART Pothole Detection System
Hao Ming Zhang, Nicholas Olson, Ugan Sivagnanenthirarajah, Andrew Olson, Jiuru Liu
The SMART Pothole Detection system is a vehicle mounted camera and embedded sensor platform that leverages server-based machine learning and image processing to visually detect potholes. The system displays the location and image of detected potholes on an accompanying mobile app and allows users to easily visualize pothole locations on roadways of interest. The SMART Pothole Detection system provides municipalities with a low cost, quick deployment road quality monitoring solution that reduces maintenance costs while making the roads safer for everyone.
Mask: A Decentralized Social Mediat
Haowen Zhan, Juejie Pan, Yuxin Jiang, Sitong Liu, Cong Zhou
Today’s social media is dominated by the service providers' centralized power. Consequently, users' privacy and freedom online are widely violated. To solve this problem, we propose to create a decentralized social media that hand privacy and autonomy back to the user community. In the proposed solution, user data is encrypted and distributed among multiple devices throughout the network, and consensus algorithms are used to ensure the data is resistant to tampering by any single entity.
Md Tashin Asheque, Kevin Abraham, Sherman Ying, Brandon Sahadeo
A 2011 United Nations report found that one-third of global food production is wasted. FoodFileSystem manages food inventory and expiration dates, to facilitate informed decision making when procuring or consuming food. FoodFileSystem provides a convenient device at the refrigerator for users to track stocked food, and a connection to a mobile application that alerts the user of foods that are expiring. The main advantages of this design is affordability over using a smart fridge, and the ability to use the system on any existing refrigerator.
Pimp My Stretcher
Will Liu, Jordan Edwards, Kelly Kong, Gilly Lovato
In recent years, stretcher wait times in hospitals worldwide have skyrocketed due to over-occupancy of hospital beds. Existing stretchers offer limited monitoring capabilities, making them ill-suited for continued care. Pimp My Stretcher is a low-cost stretcher attachment that allows hospital staff to remotely monitor patient vitals in real-time. Sensor data is collected and transmitted to an application that alerts hospital staff during emergencies and can analyze historical data for insight on trends in patient health.
Dulwin Jayalath, Justin Trudell, Yisen Wang, Vishal Kuo
The sharing of patient history and other medical records across medical institutions is a slow and non-standard process, often involving proprietary, non-electronic methods of data transfer. In 2015, despite most U.S. hospitals already possessing digitized records, only 30% supported electronic record sharing. The objective of this project is to design a system that provides a secure, accessible storage location for patient data. medFS achieves this goal by leveraging distributed file store technologies paired with an access control system and end-to-end file encryption.
Interio[V]R: A web-based, VR-enabled interior design platform
Erik Desjardins, Douglas Wong, Kyle Yeung, Ernest Wong, Matthew Cyr
Interio[V]R is a virtual-reality (VR) enabled web platform for users to plan and design interior spaces, giving users complete autonomy in choosing structural specifications, furniture, and decorations. Its VR interface provides an immersive experience that shows the true scale of objects in the room and is more natural to navigate through than traditional bird's eye 3D rendering. Intuitive drag-and-drop interactions facilitate easy placement of furniture without leaving the VR environment or switching to different display modes of furniture without leaving the VR environment or switching to different display modes.
Kura: An end-to-end encrypted, synchronized, global file system
Franz-Julius Chen, Ankit Bahl, Irfan Sharif, Arham Ahmed
Users expect to seamlessly transition from using one device to another and have their application sessions carry over. However, storage is usually anchored to specific devices, is inaccessible across devices, and limited to device specifications. Having synchronized data across devices makes for a more fluid user experience and enables new workflows using data not possible before. Also, data privacy and ownership have become a growing concern over the years. It would be useful to have tools to create systems that default to storing data securely.
RUOK: Emergency Detection System
Yifan Liu, Xiaoyou Zhang, Hui Qiao, Zijian Huang
According to the Centers for Disease Control and Prevention, 234,000 people suffer from traumatic brain injuries and bone fractures caused by slipping and falling in the bathroom in US each year. RUOK is an integrated system that can automatically detect fall or unconsciousness within a small space such as washroom and immediately seeking for help. It also provides assistances when user call for help under emergency situation, and has the potential social impact of saving lives upon emergency situations.
Aditya Joshi, Munaz Rahman, Santanu Sarker, Tanvir Pathan
Travel journal is an application that is built based on a camera that periodically records throughout the user's day, detecting and storing key moments and filtering out the rest. Leveraging machine learning, computer vision algorithms and location data to identify moments of interest, the user’s day is summarized in an e-journal to be viewed at a later time. The advantage to this design is that it is lightweight, fully automated and requires minimal action from the user.
Michel Jiang, Zhaorui Luo, Nathan Liu, Sung Mihn Lee, Ali Elwazir
Noise pollution affects both one's health and remains an unsolved problem in many urban areas that can cause hypertension, high stress levels, sleep disturbances, hearing loss, and other harmful effects. For example, professionals working in loud environments (such as construction) must wear hearing protection which hinders their ability to communicate with colleagues. The objective of this project is to design a pair of headphones that dampens or cancels surrounding sounds while selectively filtering and amplifying desired voices.
Graham McColl, Liam Benson, Henry Saffrey, Evelyn Raftery, Peter Hyungrok Sung
One of the most common issues consumers express with regards to grocery shopping is the large wait times associated with queueing for tills. This project’s objective is to design a shopping cart that keeps track of items that are placed in it, enabling the customer to leave the store quickly without an extensive checkout process. This design aims to be simpler to use for the customer, and much cheaper to implement for the grocery store over current alternatives like self-checkout stations.
Himanshu Shrivastava, Smit Patel, Ian Emdin, Hogenbirk, Anirudh Prasad, Akila Jayawardane
Many injuries, sports-related or otherwise, occur in the foot. Avoiding the habits that lead to these injuries can be challenging without a medical professional. Collecting data on the biomechanics of the foot can help identify injury-related actions, allowing the user to correct them. The objective of this project is to design an insole that collects sensor data to analyze foot motion and impact. SoleSense uses pressure and motion sensors to visualize pressure and movement via a software application, without the need for expensive professional equipment.
Kathleen Neblett, Tong Li, Eashaan Tiwari, Bryanna O’Neill
VR Notes is a tool that allows users to view video in virtual reality while having the ability to write and save notes on a virtual pen and pad. The user holds a physical pen and pad and uses a virtual reality headset with a mobile phone to view the virtual video and notepad. The advantage to using the virtual reality setup on a mobile phone application over a desktop application is the advantage of portability and ease of access.
MMA - Mike's Match Analysis
Xu Qian Li, Zi Ao Jiang, Shaheryar Ali, Nuo Chen, James Fong
Judges’ decisions are always a controversial part of competitive combat sports. The results can often be biased and/or in extreme cases, manipulated. The objective of this project is to use image recognition and deep learning on top of live boxing footage to determine various statistics or attributes of the fight, such as speed, strength and accuracy of blows, or recognize fouls. The end goal is to estimate a fair score or decision for the match to assist the human judges in real time.
Volteon: The Electric Vehicle Charger Smart Switch
Yi Ming Zhao, Natalie Josephine Soscia, Rhiannon D'Silva, YuSheng Mei, Masseullah Adel
Within recent years, interest in electric vehicles has rapidly risen causing an issue for utility providers as distribution grids are not built to accommodate huge spikes in power demand. Volteon is a system that monitors and controls electric vehicle charging across a local grid to prevent overloading. The Wi-Fi controlled, add-on charger modules work in conjunction with each other and a central server to prioritize power distribution to each vehicle as per the customer’s needs. The result is a cost-effective solution for optimal charging efficiency.
Felix: A Chatbot for Mental Health
Jenna Deng, Dayeong Lee, Hannah Zhang, Jeffrey Mew
Negative stigmas and long wait times makes it challenging for people suffering from mental illnesses to obtain treatment. The objective of our project is to design a chatbot that comforts mental health patients in their time of need by using natural language processing to interpret the user’s input messages and formulate appropriate responses. The advantage of Felix is that users who are suffering from a mental health crisis can access this resource immediately, instead of having to wait potentially days or weeks for a counselor.
Shobhit Chaudhry, Rakibul Hassan, Rachit Shah, Jayant Tiwari, Christopher Vendryes
WattEx is a blockchain smart grid system which aims to motivate renewable energy generation while reducing strain on our ageing infrastructure. WattEx controllers autonomously trade energy with their local community and consistently provide this data to utilities in order to optimize costs for the network and balance load on the grid. The main advantage of WattEx over other proposed smart grid systems is that it can be built on existing infrastructure, and it leverages the security and durability of blockchain.
Raeesa Ashique, Anam Zain, Zahaan Khan, Mohammed Abdul Qadir Hashmi
Loss of independence can be very difficult for victims of quadriplegia. Allowing for freedom of movement is a very simple, effective way of providing this independence without complicated assistive technology or AI algorithms. The advantage of this product over current TDS (Tongue Drive System) systems is the simplicity of its sensing technology, making it more cost efficient. Additionally, it can be integrated as a controller with an existing motorized wheelchair, for a user who prefers the ergonomic design of his/her current product.
Blakelock Brown, Selina Li, Badar Fatima Mustafa, Almeet Bhullar, Abdulrahman Abdulghafar
Computerized systems in today’s vehicles has led to a prominence in vehicle hacking enabling hackers to access electrical control units such as the brakes. These units communicate via the controller area network (CAN) bus. The objective of this project is to design a real-time system using a physically unclonable function (PUF) to prevent unauthorized access to safety-critical components of vehicles limiting access to a CAN bus. The main advantage of this design is that it uses on-chip encryption which is essentially impossible to duplicate.
Terrainer: Marathon Route Simulator
Bjorn Hanks, Kate Lim, Bilal Majeed, Spencer Delcore
Terrainer makes training for a marathon more accessible, by automatically adjusting the incline and decline of the treadmill depending on the marathon track's slope. It enhances training accessibility by mimicking the marathon routes that are inaccessible due to weather conditions or distance, such as a marathon in another country. Terrainer provides a mobile application that takes care of route selection, Bluetooth communication, and user data storage.
Lavan Pathmakumar, Celina Chang, Abilas Sathiyanesan, Rashmi Umashankar, Bernard Lau
The GuideMe Glove incorporates the concept of echolocation to provide an affordable and intuitive experience to assist in day-to-day navigation for visually impaired users. It comprises of a microcontroller, ultrasonic sensors, and vibration motors embedded on a glove. The system detects the distance between the user and an object using ultrasonic pulses and sends vibrotactile feedback of appropriate strength depending on the distance. The GuideMe Glove is provided in a wearable form factor and uses multiple vibration motors to give more precise navigation.
Testify: Collaborative Platform for Facilitating Technical Interviews
Jason Woo, Jae Han Chung, Yuan Kun Cai, Ansley Peduru
Testify is a service that encapsulates the end to end hiring process in a technical interview. This web-based platform provides a centralized medium for facilitating collaborative discussions between interviewer and candidates. The main advantage of this project over other alternatives is that it reduces the work required to keep track of candidates who have written take home tests. This process aims to exterminate the practice of regurgitation of memorized solutions to standardized interview problems.
Dong-Hyun Kim, William Liu, Jun Yi Zhao, Samantha Marcano
Billions of passengers travel by rail and this number is increasing. VIA Rail Canada has seen ridership increases of 8-10%. Consequently, navigating through major stations has become increasingly time consuming. Our objective is to provide a smart, automated system to validate a passenger’s status while in the station. Implementation of this system involves signal transmitters, biometric technology and embedded systems. The main advantage of this design over major alternatives is its ability to precisely authenticate and validate travellers in multiple layers, while reducing manual processes.
Mapping Indoor Spaces with Autonomous Drones
Ali Hirani, Jonathon Ellis, Ashwin Krishnan, Luke Oulahen, Jin Rui Sun
This project’s objective is to have a drone navigate autonomously around an unknown indoor environment to generate a layout of the space. The drone is directed safely through its environment using obstacle detection and path planning. As the drone navigates through its environment it performs simultaneous localization and mapping to construct a map of the space in real-time. The various algorithms are informed by data collected using depth sensors mounted on the drone.
Each American driver spends $97 extra each year on parking fees; we propose a system to address this issue. The objective of this project is to create a toll system for a parking area that identifies a car based on its license plate, charges an individual for the amount of time their vehicle was parked through an online payment system.
Jerry Y. Zhang, William Schoen, Ruhan Kuang, Mostofa Akhter, Ryan Tryhorn, Runze Yuan
Nowadays, people lose several work-hours per week due to attentional issues. Often, individuals are not cognizant of when they have become inattentive or distracted. The objective of the project is to monitor the individual to detect when the person’s attention has been distracted and through haptic feedback help to refocus. The positive aspect of this will be that one will be externally reminded to focus, and it may help reduce individuals’ dependence on powerful and expensive stimulant medications which often cause side-effects.
AutoSwap: Automated Battery Swapping Device
During November 2017, global cumulative sales of electric vehicles reached 3 million units. Unfortunately, existing battery technology limits the range of such cars, and consumers are required to charge their batteries frequently. The process of charging the battery is time consuming, especially when compared to traditional gasoline refuelling. The objective of this project is to tackle the lengthy charging process by designing an automated battery swapping solution.
Time Synchronization Framework for autonomous vehicle
Licheng Zhang, Joshua Catral, Peili Chen, Siyi Yin, Xinrui Zheng
Nowadays, autonomous vehicles are equipped with more and more sensors, such as stereo-cameras, LiDAR, radar, and they all work at different sampling rates. One of the biggest challenges is to fuse those data inputs from all sensors into a complete and real-time picture of the surroundings. The objective of this project is to design a hardware-based time synchronization closed-loop control system that maximizes the sampling rate of each device while analyzing the data input in real-time to provide feedback to adjust each sensors’ sampling rate.
Soil Detectron: The Portable Soil Nutrient Detector
Ankur Mangal, Kshitij Jain, Saahiti Annamneedi, Gurmeet Johal, Alimpana Mahalanabis
The Soil Detectron is a system that informs the amatuer home gardener about the condition of their soil and provide recommendations for suitable crops within those conditions. The system involves a portable, lightweight device that is placed in a sample of soil to record readings on temperature, pH, water moisture and macronutrients, and a mobile app that monitors growth and presents recommendations. The advantage of this solution is that it employs a portable device that is able to provide instantaneous, streamlined recommendations for specific plants.
Brandon Yee, Xiyue Wang, Linyi Chen, Junru Chen, Xuan Ren
This project involves designing a 3D scanner that constructs a 3D model of any physical object. Sensors will collect data from an object, which will be rotated and vertically displaced with respect to sensors. By drawing on linear algebra and geometric analysis, the scanned data can be merged into a 3D model. This project incorporates knowledge from embedded development, robotics, mathematics, image processing, and machine learning. The project will explore the use of multiple sensor types such as lasers, IR, and cameras.
Martin Lai, Cedric Fadel, Saad Rahimi, Jeffrey Peidong Huang, Goutam Tamvada
Owl is an accessible system that allows visually impaired people to read printed and electronic text. It uses a mobile application to which text can be supplied either from a digital format or a photograph. The text is discerned using image processing. This application communicates wirelessly with a portable refreshable braille cell. The main advantage of this design over alternatives is that it provides greater interoperability with greater portability and at a significantly lower price point.
Seeded Melody Music Generation Application
Ethan Tang, David Cheung, Taylor Chen, Jason Fok, Chen Han
Composing and transcribing music is a difficult and time-consuming process that requires significant knowledge in music theory. Because different musical genres possess unique conventions and techniques for composition, it is difficult for a musician to learn the ability to compose music in a genre they are unfamiliar with. The objective of this project is to create a solution capable of re-interpreting piano Musical Instrument Digital Interface (MIDI) files of a musical genre into a different genre through the use of deep learning techniques and artificial intelligence algorithms.
Andrew Hong, Jack Mo, Kayleigh McNeil, Aaron Lam, Tu Trung Nguyen
The Audio-Image Encoder is an application of symbology which maps short audio files to images. It uses a combination of audio data compression techniques, color symbology mapping, and audio filtering concepts to address the limited shortfalls of current applications of symbology. The advantages of this design is that the Audio-Image Encoder provides an effective method of sharing audio data in an elegant image form.
Nakash Ali Babwany, Branko Mostic, Zak Devins, Sidhant Mishra
Skipify is a web and mobile application which allows users to check the wait times at various establishments. By periodically analyzing live video feeds of lineups, using machine learning algorithms, Skipify is able to estimate wait times in real time. This allows users to make an informed decision. The main advantage of this design over current alternatives is that it allows the delivery of concrete, real-time information to the user instead of a reliance on estimates generated using data from the past.
Electrostatic Audio Visualizer
Conventional electrostatic speakers are known for their high fidelity sound reproduction. However, they lack the ability to produce large amounts of bass due to the nature of their design. The objective of this project is twofold; the first is play high frequencies through the electrostatic speaker, while playing bass through a woofer to provide a more desirable sound profile. The second is to enhance the user experience by adding a visual response tailored to the audio being played to add another dimension to the listening experience that other speakers on the market fail to.
Oliver Ng, Mudassir Chughtai, Victor Chan, Adib Shalakeen, Shaan Mehta
The IoT devices market is an emerging market consisting of “smart” devices capable of wirelessly communicating with one another. As more manufacturers enter the market, they introduce more control interfaces that a user needs to interact with. This can become increasingly counter-productive to the ease-of-use of IoT devices. C* can operate as a translator between user inputs and other IoT devices by providing an interface that can emit any given signal and allow a user to map this signal to commands for other IoT devices.
Autonomous Sports Tracking Feed
Pranay Ranjan, Louis Coste, Prasaanth Sridharan, Dale Martin, Anthony Go
Human-controlled camera systems do not have the dynamic response rate needed to track fast-paced sports with a high degree of efficiency. The "Autonomous Sports Tracking Feed" project uses a combination of computer vision and computational intelligence algorithms for determining object movement and translating it into camera tracking with greater precision than a human. The main advantage of the autonomous sports tracking system is that its physical size, strength, endurance, and mobility are improved over a human operator.
Identity: Automated Government ID Verification
Richard Lee, Jason Huang, Samger Li, An Qi Zhang
Identity is a service that allows customers to upload a photo of their government identity documents and their face, then determine if that customer is fraudulent or not. Image processing will be used to extract the document from the photo while machine learning will be used to extract text, compare the customer’s face to the document’s face, and detect unique features not present in fake documents. The advantage of this design is that it automates the manual review process in existing solutions for this space.
SmartCell: Modular Battery Charging System
Erica Kim, Kristie Zhou, Angelica Sabandal, Melissa Tedesco
SmartCell is an advanced modular battery charging system that allows for individual charging of battery cells in series. Through sensors measuring open circuit voltage, current, and temperature, charging is controlled to ensure cell balancing, and State of Charge is estimated. The advantage of this design is that SmartCell provides higher reliability, as faulty cells or damaged charger units are identified and charging can easily continue in other cells, and cell balancing is achieved without additional circuitry.
Zunair Syed, Dip Patel, Brian Kong, Dishantkumar Patel, Subhadra Upadhyay
The largest problem with fans today is the wasted airflow in unoccupied space, which requires constant user adjustment to fix. The goal of this project is to provide an affordable, 360 degree automated tracking cooling system. The TrackMe Fan provides the tracked user constant airflow with fully automated tracking, wireless mode adjustment, and power efficiency optimization through presence detection. The system redirects its airflow to the tracked user using a microcontroller powered motor. This design enhances fan operation without sacrificing portability and robustness.
Daren Wang, Yuanzhen Lu, Ziyue Wang, Yun-Ting Tsai, Xin Yun Lu
In the United States 37 percent of households own a dog. In a large portion of these homes, owners are not able to play with their pets during working hours. Long periods of being alone can increase the stress level felt by a dog due to lack of active interaction. The objective of Rover is to design an interactive pet toy that can operate in autonomous or remote controlled modes to keep dogs engaged.
InSight: Head-Up Display Visor
Huangshan Xu, Daniel Chen, Eddy Yeung, Brian Luo, Brian Engio
InSight is a motorcycle helmet with a built-in heads-up display that shows information such as speed, bearing, and map directions. InSight aims to eliminate the need for a dashboard or phone to improve the user’s situational awareness on the road. The helmet provides features such as a helmet camera, Bluetooth pairing, and customizability. The advantages of InSight are that it is a cheaper, robust, integrated alternative to the current solutions on the market, which are attachments to normal motorcycle helmets.
Conghao He, Yuxin Shen, Lingyi Li, Yiran Tian
CogniStore is an intelligent device that turns a normal refrigerator into smart refrigerator. It detects, recognizes and registers new food item being stored. Manual input is also enabled for users when the system encounters a new item and learns about it. Modern mobile and web application allow users to monitor the storage at anytime and anywhere. Notifications will be sent as items’ expiration are approaching, thus avoiding waste. CogniStore is designed to solve food waste problem at an affordable price.
Pragash Sivasundaram, Minghao Ji, Tong Su, Timothy Chee Tim Mui, Yifei Li
NN2RTL is a compiler that accepts a neural network model, in TensorFlow format, and generates a hardware representation to be implemented on a FPGA. This reduces hundreds of hours of manual hardware design to a few seconds, making custom hardware acceleration more accessible to developers and allows for neural networks to be used in power limited devices.
Stabilitics: A Tremor Detection and Mitigation System
Michel Jiang, Zhaorui Luo, Nathan Liu, Sung Mihn Lee, Ali Elwazir
Pradip Upadhyaya, Shruti Dembla, Neel Talati, Chathun Kurera
The human hand plays a pivotal role in allowing humans to effectively carry out day-to-day activities. Unfortunately, more than 12 million people in North America alone suffer from tremors. Stabilitics is a wearable device that alleviates hand tremors by stimulating tendons non-invasively using mechanical actuators and a real time predictive software model. In comparison to current task specific alternatives in the market, Stabilitics employs an active approach in solving the broader problem of reducing hand tremors for all activities.
DualSense - Obstacle Detection for the Visually Impaired
Zain Jafri, Ahsan Shahid, Huagang Zhang, Syed Haseeb
DualSense is a wearable ankle device utilizing ultrasound to detect obstacles, and guide users through audio and haptic feedback. The device acts as a transceiver which uses wave phenomena and signal processing to identify information such as the distance obstructions in the user’s path. Digital logic is used to decode information from received signals and inform the user of potential obstructions. The device is designed to be wearable, light-weight, and non-obstructive.
Meep: ASL For Everyone
Marwan Ahmed, Kray Cuellar, Jazel Canseco, Zarwan Hashem
The Canadian Association of the Deaf estimates that there are roughly 357,000 deaf Canadians and 3.21 million who are hard of hearing. The usage of American Sign Language (ASL) outside the deaf community remains limited, leading to barriers in communication. The goal of this project is to bridge the gap between the deaf and hearing communities by designing a platform that gamifies learning ASL. Meep features exercises of varying difficulties and encourages competition between peers, making ASL more engaging for all ages.
Rebecca Pang, Liz Lozinsky, Suwen Li, Jovina Gananam
Existing item tracking products are used after the user loses an item and relies on smartphone apps. Lynx is a collection of up to seven interconnected wearable devices with the master device notifying the user before tracked valuables are out of range. The master device uses LEDs, sound, and haptic technology to alert when an item, with the slave device attached, leaves a set proximity. The Lynx system is energy efficient and can be recharged wirelessly. Device characteristics can be edited through a smartphone app.
AI Calorie Measurement
Neville Chan, Myeonghun Kim, Adekola Fakorede, Zichen Wang
AI Calorie Measurement is an Android application that allows users to take a picture of food and record the amount of calories. It can recognize the captured object and calculate the size of it to get its calorie amount. Users can use this application to keep track calorie intake in any places. By providing people with a more convenient way to collect dietary information, it will be easier to start and maintain a healthy lifestyle.
Jonathan Crowther, Matthew Fletcher, Tristan Gill, Donny Li, Sergei Ten
The objective of our project is to design an interface to keep track of what products are in the fridge, and when they get added or removed. The adding of items is handled through a barcode reader and a touch screen, allowing users to input their groceries. The contents of the fridge are then monitored using weight sensors, to allow automatic removal and tracking. All of this will provide users with an easy to use and affordable smart fridge.
DeckD - Concussion Analysis Toolkit
Zu Ming Zhang, Emily Hilbig, Daniel Robertson, Cassandra Hilbig
DeckD is a toolkit for healthcare and sports medicine practitioners which provides objective and accurate data to aid in the proper diagnosis of concussions. It uses augmented reality glasses along with a pupil tracking camera to determine performance related to cognitive and coordination tests and visualizes this information in an application for the practitioner. The main advantage of this design is the objectivity of the results and the low cost of the pupil tracking component compared to existing solutions.
Wearable Gym Exercise Tracker
Sai Rijal, Vlad Silin, Waleed Chaudhry, Daniel Kim
The market of wearable devices is large and growing, with 310 million units of wearable devices sold in 2017. However, the capabilities of these devices are often restricted to measuring general levels of activity, rather than keeping track of the number and quality of specific performed exercises. The objective of this project is to design a set of wearable bands that accurately monitor movement and use the data to log the user’s performance as well as to provide them with feedback on their exercise form.
Real-Time Detection and Tracking of Unmanned Aerial Vehicles
Philippe Vibien, Alex Santamaura, Shashwat Pandey, Austin Bianchini
With the increased prevalence of consumer drones and other small Unmanned Aerial Vehicles (UAVs), the task of securing protected airspace is becoming increasingly challenging. The goal of this project is to develop a system capable of detecting and tracking the location of UAVs using Electromagnetic Sensing. This is accomplished with an array of Software Defined Radios programmed to search for intentional and unintentional UAV radio emissions. Because Electromagnetic Sensing does not transmit during operation, this system is optimal for transmit sensitive environments such as airports.
WATonomous Global Mapping
Tanatorn Nantawinyoo, Tony Wu, Wook Bum Kwon, Jiahong Liu
The Society of Automotive Engineers has put forth AutoDrive, a challenge which aims to provide pathing for autonomous vehicles. The University of Waterloo has been selected to participate in the event, where competitors are to develop features for an autonomous vehicle. One critical challenge incorporates a pathfinding system to parse natural language address requests in real time. The objective of this project entails mimicking existing mapping software (such as Google Maps), and providing accurate destinations for a vehicle to follow.
Yushi Cao, Seunghee Han (Lisa), S. K. Li (Wilson), Negin Karbasi, Alex Tran
The task of a Ball boy/girl is to retrieve tennis balls, a traditionally time consuming and manual task. The objective of this project is to design an automatic ball collecting robot. The project involves applying machine vision, image processing, kinematics and embedded systems. A mechanical device will perform the collecting action. The advantage of automatic ball retrieval is that it performs an otherwise manual task, susceptible to human error. This system will make it possible to autonomously collect tennis balls at a reasonable cost.
Sensor Fusion: Mapping for Autonomous Vehicles
Ji Gao, Joe Song, Zhenyu Liu, Zipei Jerry Shi
The sensor fusion software platform combines input data from camera, lidar, and radar to generate an image of the vehicle’s surroundings in real time. Input data collected the sensors calibrated and combined to project a measured 3D point in space onto a 2D frame of reference. The corresponding points on the screen are assigned a class and depth to identify the objects, and their respective distances from the vehicle. The advantage to fusing multiple sensor input types is that it can improve accuracy of the mapping to ±5cm.
FanMe – Smart Fan Device
Wilson Tsang, Josh Young Jin Kim, Daniel Maikawa, Andrew Lay, Jiaming Wu
FanMe is a smart fan which automatically detects and tracks a user, and upon the user’s selection provides either cooling or heating capabilities. The advantage of FanMe over a traditional fan is that it does not require the user to manually fix the fan to ensure that air is constantly flowing in their desired direction. Lastly, since FanMe provides both cooling or heating capabilities, it can be utilized seasonally and serve as a permanent household fixture.
Low Voltage DC Microgrid
Keith Yan Kit Wong, Peng Chen, Ya Xi Kitty, Zhang, Jessica Youjia Lin
Currently, DC power generated by solar panels is converted to AC in order to feed into power delivery infrastructure. The objective of this project is to design a low voltage DC microgrid, independent of the existing AC distribution system in buildings. This is achieved by controlling the power flow to and from a battery pack with respect to the variable power generation from the photovoltaic cells and the load demand. Incorporating a low voltage DC microgrid increases overall system power efficiency, reducing energy costs.
WattchDog: A Smart Energy Outlet Monitoring System
Eric James Wong, Shajesh Jegatheeswaran, Nova Alam, Joseph Nguyen, Abdallah Nabil Arar
Devices plugged in the average Canadian home spend 75% of the time drawing standby power throughout the day and accounts for up to 10% of the home owner's hydro bill. The objective of this project is to design a Smart Energy Outlet Monitoring System that identifies trends in energy usage of individual appliances around the home and cut power draw from loads on standby. The device plugs between the appliance and outlet and controls power flow via a wirelessly controlled relay. Measurements are processed remotely, and detailed analytics are available to be viewed online anywhere.
Burning Rubber: Teaching Autonomous Vehicles Using Reinforcement Learning in a Racing Game
Samuel Yigzaw, Matthew Dias, Justin Sanchez, Devansh Malik, Hong Phuc Dao
This project uses reinforcement learning to improve autonomous vehicle navigation through training in a virtual racing game. The vehicle is trained by repeatedly being applied to different simulated driving scenarios. The path planning system is then validated on our testing setup, which includes an industrial conveyor belt and custom-made radio-controlled (RC) cars. An advantage of this system over existing alternatives is safety, as testing new algorithms for self-driving vehicles is less dangerous using simulations and miniature RC cars over actual vehicles on the road.
Muikku: A Real Time Photography Assistant
Mitali Chandna, Benson Lau, Varjitt Jeeva
Muikku is an interactive photography assistant which provides real-time feedback on the current composition of the picture from the camera’s viewfinder. It analyzes the image feed from the camera to identify composition elements like lines, patterns, and structures, and gives the user positioning changes to better satisfy rules of composition based on the identified elements. The advantage of this is that Muikku allows users to learn photography theory and application, without taking expensive and time-consuming formal classes, while still rendering a professional picture.
Heads Up Display-Windshield
Sagar Aryal, Donnel Varkey George, Hussain Al-Homedawy, Tahmid Zawad, Daniel Song
This project is a heads up display-windshield that gives motor-vehicle drivers access to and control of information from external devices without jeopardizing their ability to drive. This design will project information from an external device to the windshield controlled by a haptic feedback control-system from the user. The heads up display windshield will prevent distracted driving, promote driver safety, redefine the purpose of the windshield while reduce the frequency of distracted driving on the road.
Mimicry - Music Aliasing Application
Learning how to play an instrument can take months to reach competency. Musical composition is a creative activity that is not highly accessible. Admittedly, digital audio workstations (DAWs) are an existing solution which can compose music digitally. However, DAWs are expensive and often do not have an intuitive user interface. The objective of this project is to create a music aliasing application. The application is functionally composed of two main phases: pitch detection and frequency spectrum substitution. Users of the application would be able to create music easily and cheaply without training.
Adaptive Trailer Assistant
Yifan Mo, Yu Yang Jia, Shiyun Han, Mochuan Ruan, Yuqing Wu
With the fast growing demand of transportation, the long chassis or trailer has been troublesome when drivers try to reverse the vehicle into parking. This design project is intended to develop an algorithm that can help drivers to park their trailers much safer and more efficiently. To be specific, this algorithm uses global sensors to oversee the vehicles to obtain surrounding environmental information and then calculates the best parking routes to further assist the drivers in the parking process.
Yucheng Liu, Tian Zhu, Jing Zhang, Zhiyi Zhang, Mingyuan Chen
S&W Shield is a smart system that transforms any existing camera system to “private photographers” for the audience. It preprocesses video feed from the camera system in the board, and upload each frame to the cloud. The cloud persists and classifies incoming data streams accordingly. Clients can request for images of interest from the cloud through the web/mobile application. The advantage of the system is that it provides an economical solution for moment capturing which is much more affordable than private photographers.
Yuchen Zhao, Yingda Wang, Shawn Sheng, Shiqi Nian
Deaf people rely heavily on vision to perceive their surroundings. However, sight alone is limited to what is within the individual’s field of view, which can result in a considerable amount of information being rejected. Deafinitely Aware is a device that aims to solve this issue by converting sounds from a person’s surroundings into visual cues to notify the user as to which direction the sound is coming from. This device will aid deaf people by allowing them to better perceive their environment.
Cerebral Palsy Communication Apparatus
Saad Kalim, Eishan Vijay, Salman Malim, Parth Sethi, Rahul Punchhi
Approximately 1 in 323 children suffer from cerebral palsy in the United States alone. A major daily hurdle for many of these patients is communication due to limited speech capabilities. The Cerebral Palsy Communication Apparatus consists of a device with a multi-button interface allowing users to scroll through a mobile application displaying letters, numbers and image icons. The advantage of this design is that a simplified button interface makes it easier for users with cerebral palsy to quickly choose letters and form words.
Offerwise: Unified Recruitment Platform
Adrian Machado, Francis Zhang, Zhicheng Wen, Gordon Szeto
Offerwise is a unified platform that allows for talent sourcing, candidate screening, and technical interviews. Employers can specify the technical skills and experience expected of a candidate, and the system will use artificial intelligence to screen candidates to find the best matches. Coding challenges, live programming interviews, and onsite feedback for candidates can be maintained on a unified source. The platform also benefits job seekers by providing a curated job board using their information such as their resume, skills, and job preferences.
Sufyan Patel, Peter Cheng, Andrew Choi, Abdul Rehman, Vasudev Mehra
Environmental Protection Agency statistics indicate that the total amount of misplaced recyclables contribute to CO2 emissions equivalent to 52 million passenger vehicles. The objective of this project is to design a smart recycling system which can differentiate between garbage and recyclables. Various sensors will be incorporated for image processing, object detection and categorization. Machine learning will allow the system to make more accurate decisions. By enforcing good recycling practices at the individual level, we hope to exponentially reduce waste management costs down the pipe.
GymHand: A Smart Wearable Fitness Assistant
George Shen, Behkish Nassirzadeh, Amish Patel, TianYu Dai, Mostafa Flaifel
Improper technique and attempts to lift too much weight are common mistakes while exercising that leads to muscle imbalance and injuries. Current wearable fitness devices typically only monitor the body during exercises and do not have responsive feedback. GymHand is a wearable device that guides the user through weight-lifting exercises to ensure proper execution. It collects data using an array of different kinds of sensors like pressure sensors and accelerometers, processes the data, and provides immediate feedback to the user.
David Hong, Sang Hwan Kim, Chang Bok Lee
Fiture provides personalized body measurement experience for online fashion shopping. It utilizes image processing algorithms and cameras on mobile devices to detect various colored shapes embedded on Form-Fitter system. The algorithm uses detected shapes to estimate lengths and contours of body parts. In combination with collected statistical data, Fiture can make accurate measurements. The measurements are then used to generate a 3D model of the measured body, allowing users to ultimately see personalized clothing merchandise while shopping online.
Monode - Flow-Based Programming
Monode allows developers to design, prototype, develop and refactor applications by connection visual elements together. Looking at a program or a set of programs as a flow of data between independent components allows reasoning about them as a whole. Flow-based programming lends itself to the type of continuously-running, event-based, large-scale software that is prevalent today, as opposed to the sequential programming style characteristic of more traditional software. It allows non-technical decision makers to better understand and even contribute to development.
Skrybe provides a low cost alternative to aerial writing, an expensive and dying art. Traditional skywriting methods are expensive as they rely on the evaporation of gallons of smoke oil off the high temperature exhausts of planes piloted by a human to create characters in the sky. Skrybe delivers a simple drone-mounted solution to display messages at altitudes less than 400 feet using an electronic smoke emitting system that can be controlled from the ground by the user themselves. Skrybe takes writing to the sky!