Emotion Robots
The illusion of ...
If we can look past the mechanistic construction and control algorithms implemented on a machine, can we project a degree of emotional states on this machine? This research aims to investigate the role that motion behaviours can play in ones interpretation of such high level notions as emotional states. A series of experiments have been undertaken that aim to quantify and qualify peoples inherent ability to project notions of intelligence and emotional states onto machines in rationalising its behaviour |
|
Experiments
See EXPERIMENTS 1 PAGE for details. (link available when experiment finished!)
See EXPERIMENTS 2 PAGE for details. (link available when experiment finished!)
As the Khepera robots are limited in their actuator capabilities (two wheels with an optional gripper) the expressive capabilities of each robot was highly constrained.
The online experiment was designed to enable a bigger pool of subjects from more diverse backgrounds to participate in the experiment. The objective was not for people to recognise the motivation for generating such motion behaviours in the robots but rather to express what they believed was the scenario for such motion behaviours.
Results have already demonstrated people’s inclination (whether consciously or not) to overlay a degree of anthropomorphic behaviour in their interpretations of the robots’ actions. While relatively few people deduced that the robots were representing the seven dwarves, some participants did define the robots motions in terms of emotional motivations which could be associated with the partifular dwarf the behaviours were to be representative of.
These experiments are also being extended further to explore other issues (such as the influence of more explicit social associations such as costumes or storytelling through the motions of interacting robots), and shall be all cross-compared with the current baseline experiments.