Capstone Design Project Winners: Industrial Engineering
|
|
The capstone design course provides students with the opportunity to solve a problem for a real-world client and normally culminates in a project showcase at the end of the term. This year, students quickly adapted to remote learning and created final project videos to virtually showcase their work.
In this issue of the newsletter we'd like to share the winning industrial engineering projects, to see the winning mechanical engineering projects check out our previous newsletter. Congratulations to our winners! We are so please to be able to showcase their innovative solutions and excellent work completed this term.
|
|
|
First Prize & The Peri Family Industrial Engineering Design Award: A Revolutionizing In-Car Conversational Recommender System
Project Team: Sean He, Tina Shen, Viola Song, Anne Zha
Client: iNAGO Inc. Supervisor: Scott Sanner
View Project Video
While conversational systems like Siri support a hands-free environment for drivers to look for a place to eat, these rule-based systems typically don’t take into account your personal preferences or provide reasonings for the displayed results. They simply perform one-time searches and therefore do not allow an interactive conversational process with you to take further feedback and requests.
This project team aimed to solve this problem by building an in-car conversational recommender system. Their goal was to create an intelligent system that knows your preferences, explains the reasons for the personalized recommendations and converses with you interactively to filter and refine the recommendations based on your current needs and preferences.
When developing their conversational recommendation system, the team went through a number of design phases. First, they built a variety of machine-learning systems to compare and select the one that best fit their goals. Next, they leveraged natural language processing techniques and explainable AI to provide the user with context for the system’s recommendations. Finally, the team integrated their back-end project with Google Dialogflow to enable the interactive feedback feature and the conversational component of the system.
The final conversational recommender system acts as a first prototype that makes personalized and explainable recommendations while allowing interactive conversations with the user. It stands out from other existing conversational assistants by enabling these novel and intelligent features. Beyond providing drivers with a convenient and enjoyable way to find restaurants, this technology represents a new wave of interactive and natural conversational assistants that could be integrated onto mobile devices and smart home systems for a much broader scope of purposes.
|
|
|
Second Prize: Detecting & Classifying Snowboarding Jumps
Project Team: Stefan Dusciuc, Sharon Ferguson, Yalene Sivaparamanantha, Meaghan Vella
Client: Canadian Sport Institute Pacific Supervisor: Timothy Chan
View Project Video
Freestyle snowboarding is a relatively new sport on the Olympic stage, and this project team was tasked with finding a way to analyze data collected by Inertial Measurement Units (IMUs) to help athletes improve their performance and gain a competitive edge. In particular, they needed to be able to detect jumps and extract performance metrics.
After some initial research the team built a machine-learning decision tree ensemble that uses time windows as the features to detect jumps. This allowed them to detect jumps with great accuracy from the mass of data they received. Next, they used a rule-based method to provide jump metrics in order to classify the type of jumps detected. The team then put it all together to provide the client with a tool where they can import data, detect jumps and export the metrics at the click of a button.
The project team successfully built a tool that detects 100% of jumps where about one-quarter of those detected were false positives and where jumps up to 720◦ were classified correctly 90% of the time. With their project’s success there is a path towards developing a full platform of performance analysis tools to help move the sport forward for future generations.
|
|
|
Third Prize: Cardiac Care Clinic Capacity Boost
Project Team: Zain Esmail, Hader Syed Shahin, James Valencia and Michael Wong
Client: North York General Hospital Supervisor: Michael Carter
View Project Video
A common problem faced by healthcare institutions is experiencing heavy patient volumes while being tightly constrained by resources. This combination often leads to extended wait times for patients and an overloaded workload for healthcare staff.
This project team was tasked with analyzing the operations of North York General Hospital’s Cardiac Care Clinic, with the goal of optimizing workflow in order to increase patient throughput by 10%. Their first step was to conduct a time-study analysis of the patient journey throughout the clinic across multiple days. With this data, the team was able to determine the clinic's bottlenecks and areas for improvements. Key findings include identifying the extreme wait times when waiting specifically for doctors, as well as determining that the Heart Function sub-clinic provided the most opportunity for improvement given its high usage.
With this information, the team was able to develop an enhanced operation model that functioned as a dynamic scenario-based solution which optimizes staff usage rates with patient occupancy. The solution lowers the average time spent by a patient at a clinic by assigning routine appointments to Nurse Practitioners to conduct the final assessment, instead of the doctor. This eliminates the dependency on doctors for a majority of appointments, and maximizes the Nurse Practitioner capabilities, which ultimately benefits staff and patients. For severe cases, the process flow will revert back to inserting doctors into the approach.
The proposed solution addresses the client’s expected 10% patient volume increase for the subsequent year by providing a means of efficiently structuring the clinic’s workflow. This same operational structure could be applied to other healthcare institutions to help manage patient growth and improve care by leveraging the role of Nurse Practitioners.
|
|
#DisplayYourPrideChallenge
|
|
|
June is Pride Month and marks another opportunity for the U of T Engineering community to come together and display our pride. Take part in the #DisplayYourPrideChallenge and be a part of our virtual celebration.
Step 1 - Fold an origami paper heart or get creative and make your own!
Step 2 - Film your heart, strike a pose or be silly, and hold your heart back up to the camera, fully covering the camera.
Step 3 - Upload your video to uofteng.ca/DYPC.
Check out an example from the Engineering Strategic Communications team here for some inspiration and then get ready to display your pride!
|
|
Latest University Communications
Keep up to date with the latest COVID-19 communications on the University of Toronto website. Includes FAQs, HR information and links to mental health resources.
|
|
Engineering Dean's Message
Stay connected with the Engineering Faculty and read updates from the Dean on the Faculty of Applied Science and Engineering website.
|
|
Engineering Undergraduate FAQs
Find frequently asked questions about COVID-19 that are specific to undergraduate students in Engineering.
|
|
|
|
