Design of an Anthropomorphic Dexterous Hand for Expressive Piano Performance

A research paper describing a data-driven, bio-inspired robotic hand designed for piano performance. The work by Yanhong Liang, last updated in April 2026, extracts kinematic primitives from professional pianist motion capture data to embed human motion priors into the mechanical design. Experimental results demonstrate the hand achieves peak joint velocities of 53.88 rad/s and successfully performs a Grade 7 piano piece.

Use Cases

• Designing bio-inspired robotic hands based on human motion primitives extracted from motion capture data.
• Benchmarking robotic hand performance for high-speed, high-force tasks like piano playing based on measured peak velocities and force profiles.
• Developing control strategies for coordinated multi-joint manipulation under dynamic constraints, as mentioned in the description.
• Studying the transfer of human motion statistics to embodied robotic intelligence for autonomous tasks.

Strengths

• Includes specific performance metrics, such as a peak joint velocity of 53.88 rad/s.
• Design is based on large-scale motion capture data from professional pianists.
• Released under a permissive CC-BY-4.0 license.

Limitations

• The dataset is a 5.0 MB PDF document; the underlying motion capture or design data is not directly provided.
• Row count and column-level documentation for any potential data are unknown.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare, author Yanhong Liang.

Collection Method

Motion capture data extraction from professional pianists and biomechanical literature review, as described in the paper.

Freshness

Last updated 2026-04-15 05:47:38; freshness should be verified.