Skip to content

EG Glove

The EG Glove is a wireless motion-capture exoskeleton glove designed for seamless teleoperation of Linkerbot dexterous hands. Built by Stonedrum Robotics, the glove provides high-fidelity tracking of human hand movements with a capture precision of 0.005°. Weighing just 190g and utilizing ESP-NOW wireless transmission (50Hz, 20m indoor range), it offers an untethered and ergonomic solution for collecting human demonstration data or controlling robotic hands in real-time.

Software Integration

The EG Glove connects to the Linkerbot SDK via the TeleoperationController class, which handles the mapping of glove telemetry to robotic hand commands.

Calibration Steps

  1. Power on the EG Glove and ensure the wireless receiver is connected to the host PC.
  2. Hold your hand in a flat, relaxed pose.
  3. Trigger the zero-calibration routine via the SDK ([VENDOR_SPEC_REQUIRED] for exact API call).
  4. Perform a full fist closure to calibrate the maximum joint limits.

Teleoperation Flow

The standard software loop for controlling a hand with the EG Glove is as follows:

  1. read_glove_frame(): This method polls the glove hardware for the latest joint angles and returns a normalized TeleoperationFrame.
  2. apply_frame(): The controller receives the frame and processes it.
  3. hand.move_joints(): The joint targets from the frame are sent to the robotic hand hardware for execution.

Note on read_glove_frame()

Currently, the read_glove_frame() method in the base SDK raises a NotImplementedError("Glove device integration is vendor-specific."). Because different labs use different receivers and serial protocols for the ESP-NOW bridge, the exact hardware reading logic must be implemented by the user or provided via a vendor-specific plugin. You must subclass TeleoperationController and implement read_glove_frame() to interface with your specific receiver hardware.