MechSE grad student wins NSF research fellowship


Griffin Sipes is the MechSE Department’s sole recipient this year of a National Science Foundation Graduate Research Fellowship.    

Written by

Griffin SipesMechanical engineering doctoral student Griffin Sipes was recently awarded a National Science Foundation Graduate Research Fellowship.  Sipes, who earned his BS in mechanical engineering from MechSE in 2021, is the department’s sole recipient this year.  

The NSF GRFP supports outstanding graduate students in NSF-supported STEM disciplines who are pursuing research-based master’s and doctoral degrees at accredited US institutions. The five-year fellowship includes three years of financial support, including an annual stipend of $34,000 and a cost of education allowance of $12,000 to the institution.

Sipes, who conducts research in Associate Professor Mariana Kersh’s Tissue Biomechanics Lab, will be working with Professor Ian Rice in the Department of Kinesiology to understand the biomechanics of wheelchair users in their home environments. The most common injuries in manual wheelchair users are tendon injuries in the rotator cuff. However, a significant challenge in attempting to understand and minimize this injury is that the real-world loading conditions of the shoulder during wheelchair propulsion have only been studied in a laboratory environment, which can’t mimic the many types of terrain that can cause different loading conditions on the shoulder.

Currently, Sipes is working on another project in Kersh’s group – studying equine exercise intervention programs during growth – by testing out a new data collection method, which will also inform his own research. Sipes said that typically, to get highly accurate 3D motion tracking data, an expensive setup of multiple infrared cameras is needed to track physical markers placed on the subject. Often, the software used to run these cameras has trouble distinguishing the markers and cannot automatically label them, requiring manual labeling.  To overcome this challenge, Sipes is using an open-source deep learning program that tracks motion from normal video data and automatically labels all of the desired markers. This will allow the team to utilize two inexpensive GoPro cameras, which also means they can capture motion data anywhere, not just in a lab. Additionally, having labeled markers from two views lets Sipes run a 3D scene reconstruction and recover the 3D motion data with the same level of accuracy as the commercially available motion capture systems, which often cost tens of thousands of dollars.

“This fellowship will allow me to translate the research that I am currently working on to wheelchair users, an underrepresented group in research. I would like to translate my current research on motion-tracking outside the lab to this population so I can better assess the real-world loads experienced by the shoulder. We are also working with the UIUC wheelchair track and field team, which is assisting in development of a wireless, wrist-mounted sensor capable of calculating important performance metrics (speed, cadence, distance, power, propulsion force, hand contact time).

“This state-of-the-art sensor will allow us to track the applied forces on the shoulder of wheelchair users during their normal, everyday activity. This, combined with the real-world motion data, will give us the most accurate loading conditions possible – and the data will be used to develop a relationship between fatigue failure rate and anisotropy of collagen fibers in the rotator cuff tendons using the diffusion tensor imaging (DTI) at the Beckman Institute,” Sipes said.

Share this story

This story was published April 25, 2022.