In a collaboration between Heriot-Watt University in Edinburgh and University of Southern Denmark we've been working to get an attention spotter to run with the EZ-B JD. The spotter which originally was developed for the iCub robot, uses an Asus Xtion 3D sensor to track a human and any objects that come close.
Here's a short video of our work:
The spotter connects to the robot via yarp and two python modules. These modules are now available on GitHub:
In the near future we plan to use this setup to study how people interact with the JD.