OpenSim
Validated musculoskeletal model
KinoIQ runs inverse kinematics on the Lai–Arnold OpenSim model, the same open-source pipeline used in published biomechanics research — applied to standard smartphone video instead of a marker array.
Accuracy varies by plane and by metric. Sagittal-plane joint angles are the most reliable; depth-axis measurements such as pelvic drop are reported for direction rather than magnitude. The app labels every metric with its measurement confidence so you know what a number will support. Validation methods and current benchmarks are available on request.
Peer-reviewed methods, clinic-ready outputs.
Gold-standard motion capture requires six-figure lab setups, trained technicians, and hours of post-processing. That infrastructure lives in research hospitals and elite performance centers — not in outpatient clinics or on the field.
KinoIQ combines markerless pose estimation with inverse kinematics on a validated OpenSim musculoskeletal model (Lai–Arnold), producing 3D joint kinematics from standard smartphone video.
The underlying approach builds on peer-reviewed, open-source biomechanics research — OpenSim and the OpenCap marker-augmentation method. The accuracy figures are our own measurements of the KinoIQ pipeline against marker-based ground truth, not results carried over from those papers. KinoIQ reports joint kinematics; it does not estimate ground reaction forces or joint moments.
Smart Coach links corrective suggestions to current PubMed literature, so recommendations reflect published evidence rather than a static exercise library.
