Everyone who has ever read a piece of science fiction literature knows that humanity is going to employ more and more robots. They will work for us everywhere – in factories, in our homes, schools and so on. However, a new study from UCL and the University of Bristol has revealed that humans prefer having robots that are more expressive and emotional, even if it comes at a price of a lower efficiency.
These results come as a surprise, because one would imagine that a robot is just a machine and its value can be judged by how well it accomplishes the tasks we assign to it. However, people look at robots in their close environment differently and they want them to be more human even if it means they make more mistakes. Scientists achieved this knowledge by conducting experiments with a humanoid assistive robot, which was programmed to help people to make an omelette.
This robot had to pass the ingredients. In some experiments robot would drop the fake eggs and sometimes it would try making amends. Scientists wanted to see if the robot would manage to retrieve the user’s trust and how people perceive a robot that makes mistakes. It turns out, most people prefer a more communicative and expressive robot, even if it makes more mistakes and the tasks take longer to complete. Users accepted robot’s apology and positively reacted to its sad facial expression, because they could see the robot knew it made a mistake. The robot was also programmed to ask if people would hire it for their kitchens and some people were afraid to hurt its feelings so either avoided the answer or lied.
Kerstin Eder, co-supervisor of the study, explained that this research “gives key insights into how communication and emotional expressions from robots can mitigate the impact of unexpected behaviour in collaborative robotics. Complementing thorough verification and validation with sound understanding of these human factors will help engineers design robotic assistants that people can trust”. People will not treat a robot like a common household item if they show human traits.
Robots are already being developed to help us around the house. Mistakes are inevitable and they will be breaking down and causing trouble. Knowing how humans react to them and what they prefer will be essential to developing robotic assistants in the near future.