PhD candidate wins best poster award at US-Korea Conference
title Heading link
MIE PhD student Seongmi Song won the best poster award at the US-Korea Conference 2019 (UKC) held in Rosemont, Illinois, over four days in August.
UKC is an international conference attended by more than 1,000 researchers including chancellors and presidents, and 146 posters were presented during the event.
Her research entitled “Metabolic Cost Estimation using Dual Unscented Kalman Filter and Gaussian Mixture Model” was performed with UIC PhD student Prakyath Kantharaju under the direction of MIE Assistant Professor Myunghee Kim in her Rehabilitation Robotics lab in Chicago.
The research looks at the performance of assistive robotic devices, which is often evaluated by metabolic cost, and used to help enhance mobility for individuals with impaired mobilities (e.g., growing elderly population). The metabolic cost is the amount of energy needed to perform a task.
“The measurement of the metabolic cost, however, is challenging, due to its slow mitochondrial dynamics and noisy in respiratory measurement,” explained Song. “As a result, it typically takes at least five minutes to obtain a reasonable estimate per testing condition. This duration requirement is difficult, especially for the target population who has limited strength.”
According to Song, researchers have focused on reduction in metabolic cost estimation duration. One group could achieve an estimated steady-state metabolic cost of about two-three minutes measured data using their dynamic model. Using the model, they conducted real-time metabolic cost estimation with steady-state cost mapping, instantaneous cost estimation, and instantaneous cost gradient search methods.
The award-winning paper proposes two algorithms, Dual Unscented Kalman Filter and Gaussian Mixture Model, to minimize the metabolic estimation time as well as provide high confidence in the measurement.
“Seongmi is a leader and at the same time, a strong team-player. She has great enthusiasm and is diligent in her work,” Kim said. “Moreover, she has many abilities, such as great presentation skills and the ability to understand highly advanced content, and integrative problem-solving ability, which is crucial for the inter-disciplinary nature of rehabilitation robotics.”
The UKC was hosted by three organizations – the Korean-American Scientists and Engineers Association, Korean Federation of Science & Technology Societies, and Korea-U.S. Science Cooperation Center – and the theme was “Smart Science, Engineering and Health for Livable Communities.” It addressed technological challenges in the era of the fourth industrial evolution in building livable communities. The conference facilitates US-Korea cooperation on developments in science, engineering, technology, industry, and entrepreneurship, and it has a leadership program for young students and professionals.
More information about the conference can be found at https://ukc.ksea.org/ukc2019.