Over the past years, human-robot interaction has been established as a useful methodology to investigate various aspects of communication. Especially the interactional functions of linguistic features can be investigated very well using human-robot interaction experiments since robots can be manipulated and controlled in ways humans cannot – without being reductionistic; that is, in contrast to other methods preserves human-robot interaction the dynamics of interaction.
We have built up an infrastructure for human-robot interaction experiments for our students to use. The approach we provide for our students is to use robots to test in controlled interactional scenarios the functions of selected linguistic features.
The infrastructure provided comprises the following resources:
- A toy robot (Robo Sapien V1);
- An iCub graphical robot simulator;
- Two Keepons to carry out studies using the method introduced by Andrist et al. 2013 and 2015;
- Two Pleo RB's;
- Three EZ-robots;
- Xbox 360 and One Kinect sensor;
- Hardware to setup Wizard of Oz studies
- An online survey tool that supports video playback (available here);
- Instructions on how to edit and manipulate speech signals using Audacity, Praat, and ProsodyPro.
- Guidance on how to create and modify utterances using text-to-speech (TTS) engines, which can be used with the simulator or other video material;
- Guidance and tutoring on how to use 'crowdsourcing' on HRI research;
- Instructions on how to do HRI research.