MiniBEE: A New Form Factor for Compact Bimanual Dexterity
Sharfin Islam, Zewen Chen, Zhanpeng He, Swapneel Bhatt, Andres Permuy, Brock Taylor, James Vickery, Zhengbin Lu, Cheng Zhang, Pedro Piacenza, Matei Ciocarlie
AI summary
Problem
Traditional bimanual robots rely on two highly articulated arms, creating large footprints, high complexity, and a restricted dexterous workspace that only covers a fraction of the system's reach.
Approach
The authors propose MiniBEE, a compact system that links two reduced-mobility arms into a single kinematic chain, using a novel kinematic dexterity metric to optimize relative gripper positioning while minimizing size and weight.
Key results
- Introduced a kinematic dexterity metric to evaluate and optimize relative gripper positioning
- Designed an 8-DOF compact configuration matching the dexterity of traditional 12-DOF systems
- Enabled self-tracked wearable data collection without external tracking or SLAM
- Demonstrated end-to-end imitation learning from wearable demonstrations for robust real-world manipulation
Why it matters
It provides a scalable, low-footprint solution for collecting high-quality bimanual demonstrations and deploying them on standard robot arms, advancing accessible and mobile robotic manipulation.
Abstract
Bimanual robot manipulators can achieve impres- sive dexterity, but typically rely on two full six- or seven-degree- of-freedom arms so that paired grippers can coordinate effec- tively. This traditional framework increases system complexity and footprint while only exploiting a fraction of the overall workspace for dexterous interaction. We introduce MiniBEE (Miniature Bimanual End-effector), a compact system in which two reduced-mobility arms (3+ DOF each) are coupled into a kinematic chain that preserves full relative positioning between grippers and enables the entirety of systems workspace to be used for dexterity. To guide our design, we formulate a kine- matic dexterity metric to evaluate different kinematic designs. The resulting system supports two complementary modes: (i) wearable kinesthetic data collection with self-tracked gripper poses, and (ii) deployment on a standard robot arm, extending dexterity across its entire workspace. We present kinematic analysis and design optimization methods for maximizing dex- terous range, and demonstrate an end-to-end pipeline in which wearable demonstrations train imitation learning policies that perform robust, real-world bimanual manipulation.