Welcome.
This is Andrea Grossi.

Built a marker-less, out-of-lab method to track a climber's 3D Center of Mass during real climbs — replacing expensive Vicon-style motion-capture rigs with a single camera and a sensorized climbing wall.

The system fuses two data streams: a neural network (MediaPipe) extracts 2D pose from monocular video, while tri-axial load cells capture the 3D force vector at each grip. A geometric model reconstructs the Center of Mass; force data fills the depth coordinate the camera cannot see.

Validated against a Vicon optoelectronic system. Final accuracy: ~1.5 cm on x/y, ~5 cm on z — competitive with lab-grade systems, usable in real gyms with children in rehabilitation.

International student challenge by Movella (Xsens) on cross-country skiing biomechanics. Athletes performed on snow wearing a full-body Xsens IMU motion-capture suit, generating high-frequency 3D kinematic data of every joint during real strides.

Our team — SkiSense Innovations — built a performance-analysis pipeline focused on technique, symmetry and efficiency, designed for athletes and coaches. Selected as finalists among teams from across Europe.

The project combined real athletes, real sensors and real performance questions — the same loop I try to bring into product work: identify a meaningful need, build something rigorous, validate it in the field.