Skip to content

GeoIK Visual Servoing

GeoIK velocity control demo

Geometric IK-based velocity control with eye-in-hand visual servoing for Franka Emika robots. Track ArUco markers in real-time and maintain a locked end-effector-to-marker pose.

Quick Start

1. Install and Run C++ Velocity Server (inside ServoBox VM)

servobox pkg-install geoik-velctrl
servobox run geoik-velctrl

2. Clone geoik-velctrl on the host PC

git clone https://github.com/kvasios/geoik-velctrl.git
cd geoik-velctrl

3. Install Python Marker Tracking (host PC)

Create environment with micromamba (or conda/mamba):

micromamba create -n marker-track python=3.10
micromamba activate marker-track
pip install -r requirements.txt

4. Run Marker Tracker (host PC)

python3 scripts/marker_track.py

Press t to start/stop tracking. The robot maintains the locked end-effector-to-marker pose as you move the marker.

Resources

  • GitHub repository: kvasios/geoik-velctrl
  • Full documentation and custom build instructions in project README