CAPSTONE DESIGN PARTICIPANTS
Sean Andersen
Yi Jiang
Ryan Pitchford
Yitian Wang
Wildlife Monitoring System
44
The Wildlife Monitoring System is a remote wireless trail camera system which autonomously monitors, identifies, and generates analytics on deer in a given area. The system utilizes custom low-power long-range wireless trail cameras to capture and transmit wildlife images to a server where the species are identified through image processing. The main advantage of this design is that it automatically identifies species, operates in remote locations, and does not require manual retrieval of trail camera images.
Owais Ahmed
Mohammed Al-Ani
Ahmed Elsaigh
Saaem Rizvi
Chiagozie Robert-Maduekwe
Arnab Saha
IntelliHelm:
A Smart Motorcycle Helmet
45
IntelliHelm is a motorcycle helmet which incorporates a multitude of distraction mitigation and safety solutions to aid motorcyclists. Information from a blind-spot detection system, a rear-view camera, and real-time navigation data are displayed on the visor in a non-intrusive manner. Furthermore, voice assistance is implemented for the user’s ease and a breathalyzer is integrated to discourage intoxicated motoring. Intellihelm’s benefits include a low-cost design and an all-encompassing suite of safety features, which are typically left out of current alternatives on the market.
Rahul Chakravarthy
Chris Ding
Jacob Kim
Leo Liu
CrowdGrip
46
CrowdGrip is a crowd-sourced data sharing platform that allows professionals and enthusiasts alike to have access to high-quality and diverse training data for their custom machine learning models. Using the CrowdGrip app, users can record and upload real world data on-the-fly. Then, software developers can use the CrowdGrip data consumption tool to download, stream, or further curate collected data. Through providing an open and public platform for aggregating training data, CrowdGrip solves the fundamental challenges of training data diversity and accessibility when creating ML models.
Samer Mohamed
Dan Rafael Timonera
Zarak Yahya
Not pictured: Chen Xu Lei, Jasper Kenichi Nakao, Anupa Isuraka Rajapaksha
Medical Conversational AI for Patient Diagnosis
47
The COVID-19 pandemic has revealed that our current medical systems lack efficiency in handling large volumes of patients. By installing kiosks in medical centres through which patients converse with an artificial intelligence (AI) to assess their symptoms and produce potential diagnoses before they interact with the doctor, patient wait times and staff loads are reduced. The main advantages of our solution over chatbots are direct integration into medical centres, direct review of diagnoses by doctors, measurement of vital signs, and accommodation of the visually impaired.
Andre Cayen
Gauri Goyal
Huixin Jin
Meng Zhao
Felix Zheng
Solit:
Emergency Solar Lighting System
48
Solit is a low-cost, modular, portable and easy-to-use solar lighting system to help health practitioners in refugee settlements and peri-slum areas in northern Uganda increase the quality of care. A battery stores energy produced during the day and a suite of sensors capture system usage data which is uploaded to the cloud for analysis and monitoring. The main advantages of this design over other alternatives are that the system optimizes solar energy harvesting at an affordable price and data is collected to improve future installations.
Mridul Ghanshala
Rohan Gupta
Farzan Nadeem
Rajan Nijjar
Zubair Waheed
Digit
49
People all around the world suffer from carpal tunnel syndrome which severely hinders their ability to use a computer. This problem extends to others with limited hand mobility, such as the elderly. The aim of this project, Digit, is to reimagine the way we use computers to make it more accessible for those who struggle to use a mouse. Digit analyzes finger movements and maps them to computer UI actions, allowing for seamless navigation at your fingertips.
Christian Feng
Jiwan Kang
Ishan Kapoor
Aswin Logan
SunFyre Sports
50
SunFyre Sports is a sports analytics application that gives users feedback/analysis based on inputted game footage. The application uses computer vision, machine learning and image/video processing to convert the video input into a digitized format that can be analyzed and studied to see where players/teams can improve. The advantage of this automated, data-driven approach is that it enables unprecedented throughput in footage analysis and can include massive amounts of data in the developed model.