In This Story
Twelve teams in the Department of Electrical and Computer Engineering (ECE) presented their senior design projects to peers, faculty, staff (and even some parents) on December 5. Topics included a multi-robot pianist, drone swarm attacks, and real-time battery monitoring.
Teams work on their project for two semesters; at the presentation, they are scored by an evaluation team comprising industry professionals, graduate students, and faculty members. The best project wins the ECE Design Award, which Tolga Sayata, a professor in the department and coordinator of the projects, said, “is like the Grammys or Oscars of ECE.”
The teams’ projects combine solid foundations in engineering and design, with a practical, useful device as an end-goal. While many focused on life sciences and health technology, each team aimed to improve and make more efficient existing technologies.
While design work is key, Sayata stressed the importance of business communication. “We put a lot of emphasis on being able to present your work to peers and outside-the-subject-area individuals. Acquiring these presentations skills is important for succeeding in their professional life.”
The Smart-Foot Scanner team: Joseph Arthur, Sawera Ashfaq, Zekarias Belachew, William Fields, Edward Fleming , and Yulieth Larobardiere.
One team brought their smart-foot scanner to demonstrate how a simple device can reveal a lot about a person’s health. Originally given a goal of making a sock to help an elderly individual who had recurring foot issues, the team instead designed an at-home system aimed at reducing doctor visits and improving foot health.
“The test subject would wake up, turn to the side of his bed, take off his sock if he slept with it on, and then step on an outline of footprints on the device (pictured),” said team member Sawera Ashfaq. Once the weight activated switches, the foot is subjected to a 15-second scan, measuring moisture, pulse, and temperature. In addition, cameras take an image of the foot. “The saturation, the temperature of the foot, and other data combined can give really good information about foot health, mainly, and then a little bit ofgeneral health as well,” said Ashfaq.
Team member Edward Fleming then designed a convolutional neural network to take pictures from the top and bottom to detect fungus or any other abnormalities.
The Tiger Muon Detection team: Arda Cobanoglu, Danie Uribe Diaz, JP McLeese III, and Kevin Nguyen
The Tiger Muon Detection team is on the lookout for muons, which are subatomic particles comprising much of the cosmic radiation that reaches the earth's surface. “Muons are the second decayed version of alpha, gamma, and beta particles. Detecting them is important regarding hardware and different satellites that are in space, and understanding how they affect those circuits,” said Danie Uribe Diaz.
JP McLeese said the team built on—and improved—current technology. “We used an existing design from MIT, but we made it significantly smaller. We created something that allows for multiple bands to detect different energy levels of muons, as well as scaling it to size that can fit in a 3.5-centimeter cube. Then we analyze findings on a web app to be able to do analysis.”
In addition to designing the device, the team worked alongside a local high school, with students who are part of a program called Cubes in Space. The team shared their work with the younger students to help them prepare for a launch next summer.
Goal-Line Technology team: Joseph Kim, Phap Nguyen, Alex Clarke, Dwayne Bass, Diego Perez Sanchez, Vily Dubon, and Luis Chica.
A third team worked on soccer goal-line technology. Currently in major soccer leagues, a system costing between $200,000 to $2 million per field is used to determine if a ball crosses a goal line; a review that happens surprisingly often. Wanting to find a cheaper method that would be widely accessible, the students created a system comprising two cameras and then a nano that is around 90 percent accurate.
“We have a system that is simple, but still capable of detecting fast-moving objects very precisely. So we wanted to simplify it and make it affordable, without losing too much accuracy,” said a team member. “We explored different algorithms, different hardware setups, and ultimately came down to relying on simple hardware but with utilizing technologies like machine learning.” The students created an algorithm that can precisely detect a soccer ball going a particular speed within a camera frame.
The system employs a machine learning model that uses a convolutional neural network to detect that a ball is actually on the field, and then the model finds the position of the ball. One team member said, “It will record video, see the goal line, map the goal line, and then it will detect the ball and where it is and how accurate that detection is. From there, we ensured that the software aspect was as precise and detailed as possible.”