Rowing Reimagined

About

Rowing Reimagined is a research project at the University of Twente's HMI lab that created an immersive VR rowing experience in the Resonite platform. The main goal was to let users practice and race together in a shared virtual world. My role was to provide technical support on two fronts: first, by keeping the project's specialized RP3 rowing machines repaired and running. Second, I developed a completely separate pose-tracking system for a master's student who needed a way to analyze rowing form without expensive sensors. This system uses computer vision to provide feedback on a rower's technique.

My Role & Technical Hurdles

One major challenge was finding a simple way to analyze a rower's form. For a master's student's project, I built a solution using Python, Mediapipe, and OpenCV. The system uses a standard camera to track the rower's body. I trained a pose model that could identify key body landmarks, calculate the angles of the rower's back, knees, and arms, and provide real-time feedback on their form and stage in the stroke. This provided a low-cost, effective way to gather biomechanical data. Separately, I also became the go-to person for hardware maintenance, troubleshooting and repairing the RP3 rowing machines to ensure the VR team's research wasn't delayed.

Results

The pose-tracking system I built became a valuable tool, offering a practical and affordable method for technique analysis that previously required much more complex equipment. It successfully demonstrated how computer vision can be applied to sports science. My hardware support for the main VR project ensured the researchers had reliable equipment, allowing them to focus on their goal of creating an immersive, multi-user rowing simulation.