Thursday, 2 February 2012

Understanding Robot Acceptance

Robots are increasingly being applied to home and healthcare settings, designed to serve as personal systems.  The concept of personal use implies that a robot may interact socially or collaboratively with users having little or no formal training. Given that the use of personal robots may be expected to become a part of people‟s everyday lives and that radical technologies, such as robots, may not be as readily accepted as incremental technology, it is critical to understand the factors that may increase acceptance and adoption.

It is generally accepted in the research community that people are willing to apply social characteristics to technology.  Humans have been shown to apply social characteristics to computers, even though the users admit that they believe these technologies do not possess actual human-like emotions, characteristics, or “selves”.  Humans have been shown to elicit social behaviors toward computers mindlessly, and treat computers as teammates and having personality, similar to human-human interaction.

However, effective and intelligent social interaction requires more than simply applying social characteristics to a robot.  The level of social intelligence needs to meet the users’ expectations for acceptance to take place. [...] According to Breazeal, when designing robots, the emphasis should not be whether people will develop a social model to understand robots.  Rather, to facilitate acceptance, it is more important that the robot adhere to the social models the humans expect.

People form quick impressions about an entity even when little information about it is available in the environment. [...] People may be prone to judge the overall characteristics of a robot by merely assessing its external features.  Thus, to ensure a successful human-robotic interaction, it has been proposed that the robot be given a form that enables people to intuitively understand its behavior.

Studies investigating preferences for the human-likeness of robots have provided mixed results.  Overall, young adults (university undergraduates) have been found to have a preference for a human-like appearance of robots.  However, as compared to others, introverts and individuals with low emotional stability are more likely to prefer mechanical looking robots.
Robot appearance not only affects preference, but also people‟s perceptions of the robot’s personality. For instance, in an examination of the relationship between robot appearance and personality, one study found that children between the age of 9 and 11 considered robots with mixed human-robot features to be friendlier than completely mechanical looking robots.  However, they judged pure human-like robots (i.e., robots modeled after human form and features) to be the most aggressive of all other robots. Robins, Dautenhahn, Boekhorst, and Billard (2004) investigated the effect of robot’s human-likeness on the level of interaction with it
by children with autism.  By measuring the duration of eye gaze, touch, imitation, and physical closeness with the robot, the authors inferred that children with autism in their initial response preferred to interact with a plain, featureless robot to a more human-like robot.

Although androids and some humanoids are modeled either after a man or a woman, many existing robots do not have an apparent male or female gender. Some studies have attempted to examine if people would attribute gender to a robot even when its appearance was not clearly indicative of one.  One such study used robot pictures to evaluate children’s perceptions of robots and found that most children between the ages of 9 and 11 assigned gender to robots, particularly male gender. Children ascribed female gender to some robots that they also rated high on primarily positive traits, such as friendliness and happiness. However, robots that were assigned male gender were associated with both positive and negative traits.  This study, therefore, recommended designing female gendered robots for children; but again this is a premature conclusion based on a single study.

[...] when the gender of a humanoid robot is clearly evident, it may influence
how people evaluate the robot. Siegel, Breazeal, and Norton (2009) found that people tend to
rate the robot of the opposite sex of themselves as more credible, trustworthy, and engaging.

Probably because the translation of Mori’s original article [on The Uncanny Valley] used “familiarity”, researchers investigating the uncanny valley have attempted to measure how familiar people find robots based on their appearances. However, it is an ambiguous construct.  Familiarity is likely to change with increased interactions.  What appears strange initially may become familiar after a few encounters. Moreover, high familiarity may not necessarily imply liking or acceptance. Similarly, low familiarity may not always imply disliking or rejection because if it does, it will mean that people do not ever like innovations or creativity.

Rosie from the Jetsons, C3P0 and R2D2 from Star Wars, and Robbie the
Robot from Forbidden Planet, are all beloved science fiction characters and they have influenced the way in which the general public thinks about robotics.  This media exposure may create preconceived expectations about robots, even for individuals who have never interacted with a robot directly.  In fact, research has found that most people have ideas or definitions of what a robot should be like. [...] Any mismatches between user expectations and a robot‟s actual capability, however, would be expected to negatively impact acceptance. Additional research is needed to explore these preconceived expectations further and to fully
understand how they impact robot acceptance.  

Excerpts from:
‘Understanding Robot Acceptance’

by Jenay M. Beer,
Akanksha Prakash,
Tracy L. Mitzner,
Wendy A. Rogers

No comments:

Post a Comment