CE Capstone - Projects
2023-2024
Chirality
Members
Diego Jerez
Ananth Pilaka
Jonathan Wilcox
Phil Wang
Yusheng Su
Description
As people’s daily lives become more and more automated, smart home devices have become more and more prevalent, yet we continue to lack a simple all-in-one solution to control all of our smart devices. This project aims to create a dedicated and simplified user process for smart home devices by providing a smartphone application and easy cloud connection for any Bluetooth or wifi devices. By connecting to the cloud, we can effectively pull and push data from your phone to your smart device and vice versa.
Resources
Sponsors
Concordia
Members
Logan Geivet
Alexander Robles
Jack Newman
Nick Pellegrin
Peter Lototsky
Description
As people’s daily lives begin to become more and more automated, smart home devices have become more and more prevalent in our everyday lives yet continue to lack a simple all in one solution to control all of your smart devices. Circuit Sages is aiming to create a dedicated and simplified user process for smart home devices by providing a smartphone application and easy cloud connection for any of your bluetooth or wifi connected devices. By connecting to the cloud we can effectively pull and push data from your phone to your smart device and vice versa.
Resources
Sponsors
Empro
Members
Elizabeth Rakotyanskaya
Jacob Woodward
Marco Mora
Neil Balch
Description
Empro is developing a new alternative to e-bikes and conversion kits. This innovation democratizes electric biking, making it accessible to millions of cyclists with minimal installation effort for an affordable price. With the batteries and motor housed on the front wheel, the installation process becomes quick and, most importantly - safe. The onboard electronics coupled with a phone application enhance the experience by enabling precise performance tracking and offers location services for added peace of mind against theft. The slick design transforms any forgotten old bike into a fun new way to enjoy one's weekends with family and friends.
Resources
Sponsors
EyeMatic
Members
Michelle Ly
Andrew Chen
Ethan Riley Nguyen
Kenya Aridomi
Marco Wong
Description
Cataract surgery is primarily affected by the rotation of the patient's eye. The rotation of the eye changes when the patient lays down for operation. In this project, through the use of machine learning, the surgical camera system can be taught how to differentiate the anatomy of the eye's anterior structure, which consists of the pupil, iris, and sclera, and detect the eye's rotation and movement. By utilizing videos of actual eye surgeries, we will be training the neural network model with tagged images. Once the model is trained, the next step involves deploying the model to our PolarFire FPGA board. The FPGA also allows for HDMI pass-through, enabling the recognition and classification of the eye from videos outputted by the computer.
Resources
Sponsors
FSAE
Members
Joshua Thomas
Aaditya Channabasappa
Alvin Liou
Henrique Da Ponte
Kane Deng
Description
The Society of Automotive Engineers (SAE) hosts a collegiate competition annually called FSAE Electric, where students from colleges across the world are tasked with building a formula-style electric vehicle that will compete in different challenges that will test its endurance, acceleration, braking, and most importantly, its accordance with technical specifications. Our capstone project is centered around collecting valuable sensor data for Gaucho Racing’s GR24 vehicle. It will provide the engineers and the driver with valuable information to improve our future vehicles and ensure our car is operating under its known tolerances. This involves creating a system of zonal nodes that collect data from a suite of sensors and transmit this data over CAN throughout the rest of the car. The data collected will be shared with our pit crew, analyzed in the cloud, inform the driver of the condition of the car, and inform our other controllers that we are operating safely.
Resources
Sponsors
Homeflow
Members
Brandon Lee
Ben Cruttenden
Harshita Gangaswamy
Julian Frank
Snehith Nayak
Description
HomeFlow is an advanced yet affordable health wearable, designed with a watch form factor and specifically tailored to the elderly. It captures relevant data such as heart rate, blood oxygen levels, body temperature, and environmental weather conditions directly from the wearer’s wrist. This allows important metrics to be collected in real time and logged continuously.The collected data is transmitted to a custom Android app via Bluetooth, where it is stored and analyzed. Here, machine learning algorithms are used to determine numerous additional health-related statistics. In medical care facilities, such as rehabilitation centers or assisted living residences, these metrics and statistics could be used to remotely monitor the condition of residents.
Resources
Sponsors
Locus
Members
Cameron Cummins
Ethan Lee
Minh Bui
Description
Oceanography researchers aim to measure turbulent flow under the ocean to learn useful weather information and predict certain events such as tornadoes and vortexes. Turbulent analysis also allows researchers to learn about the mapping of solids under the ocean. Currently, means of measuring turbulence are acoustic velocimeters that are expensive and easily break. The LOCUS team at UCSB aims to explore alternative ways of measuring turbulence that are cheaper and more accessible, to make oceanography and turbulence study more available. As such, the study and understanding of oceanography in the community will be further developed, as it benefits all of us.
Resources
Sponsors
Lumirail
Members
Jake Greenbaum
Chris Fisher
Jack Shoemaker
Sam Ng
Zach Richards
Description
Our project aims to create a captivating LEGO art piece depicting the Massachusetts Bay Transportation Authority (MBTA) map, with added features to provide real-time information of the subway system. Through several LEDs spread throughout the map, this project visually represents the precise locations of trains within the MBTA network. The companion mobile app will empower users to seamlessly connect the artwork to the internet, facilitating the acquisition of real-time train data and unlocking a host of personalized map features. These capabilities encompass tracking a specific train’s journey within the system and offering clear guidance on navigating from one station to another, enhancing the overall interactivity and functionality of this creative masterpiece.
Resources
Sponsors
PenGUI
Members
Torin Schlunk
Ayush Shah
Rodney Li
Rylan Pow
Simon Lai
Description
PenGUI connects to Pravieum’s VSCEL laser dev kit and allows researchers to tune the laser to their specific needs. Our project aims to replace an online web based GUI with a touch screen offline GUI to control the laser. The touchscreen is a reTerminal which is powered by a Raspberry Pi compute module. Additionally, we will have an adapter board which the reTerminal will connect to so it can properly communicate with the laser. This adapter board will allow for the deprecation of a Red Pitaya microcontroller currently used to control the laser which is significantly more expensive than a reTerminal. Overall, our project's goal is to deliver a better interface for laser tuning at a more affordable price.
Resources
Sponsors
Pete
Members
Anoushka Sawant
Aaron Sin
Frank Yao
Sophie Guan
Spencer Tang
Description
P.E.T.E. (Procedure Execution Tracking Engine) is a tool designed to help astronauts track procedures without needing ground support. To demonstrate this proof of concept, we chose removing the bottom bracket of a bike as P.E.T.E.'s task. We have two methods planned to help evaluate the accurate completion of each step. First, we have a computer vision model that can identify the different tools necessary for the procedure. The model uses YOLOV7 and is trained on a dataset consisting of six tools. Second, we interface with sensors placed on tools so P.E.T.E can track the tools' orientation and movement. The object-detection model is run on NVIDIA Jetson Orin, and uses live feed from wide-angle and close-range cameras as input. A GUI displayed on a monitor allows the user to interact with P.E.T.E.
Resources
Sponsors
S.E.A Boat
Members
Baron Young
Cameron Barrett
Dhruv Aggarwal
Maxwell Jung
Wenjin Li
Description
Military reconnaissance demands an unmanned vessel capable of autonomously navigating waterways, gathering intelligence, and executing reconnaissance missions without human intervention. Simultaneously, the vessel's adaptability for coastal monitoring introduces the need for a technology that transcends the limitations of existing airborne solutions. The "Kayak Drone" is designed to be a lightweight and maneuverable vessel that enables remote exploration and mapping of the littoral zone. Lacking dependence on classical wireless technologies such as cellular, it is designed to operate in regions where self-sufficiency is the norm. It will be capable of remote operation as well as autonomous mission execution.
Resources
Sponsors
United Sensors
Members
Edison Chen
Ethan Dat Nguyen
Hector Moreno
Shabeeb Reza
Tim Qin
Description
This project is concerned with developing robustness with respect to sensor failures during autonomous drone flights. Our team is developing a custom PCB to be mounted on the drone. This PCB will feature multiple sensors of the same type to prove redundancy is possible and will be programmed using custom drivers. These drivers will be published to the open-source Arducopter library once they work as intended, and as our final goal, we would like to demo a flight using a quadcopter that features sensor failure mid-flight.