Research & Development

Movement Intelligence
for Bipedal Agents

Research from Taiji Motion on movement quality, measurement frameworks, and the deployment requirements for humanoid robots in human-facing environments. Papers are offered openly under CC BY 4.0 for use by the research community.


White Paper 01

Biomechanics-Driven Foundation Models: Bridging the Proprioceptive Gap

Haiming Chen  ·  Founder & Systems Engineer, Taiji Motion

Current humanoid agents face a "Proprioceptive Gap" — the inability to move with proactive stability. This research proposes a framework for encoding Taiji's biomechanical principles as movement primitives, shifting robotic control from reactive sensing to master-level proactive balance.


In Development

Measurement Specification — Position Paper 02

A companion technical document that operationalizes the Deployment-Grade Motion framework into specific, reproducible measurement criteria — sensor requirements, threshold definitions, inter-rater reliability standards, and domain-specific benchmarks across Performance, Communication, and Integration layers. The first section (Smoothness, Performance Layer 4.1) is in active drafting.

Interested in contributing to the Measurement Specification?  Get in touch →