A Case Study Interview with the University of Waterloo
With more than 42,000+ students attending annually, and with the largest engineering school in Canada, the University of Waterloo is #1 in Canada for experiential learning and employer-student connections. Waterloo have been using MiRo for various research projects, and the most recent published paper titled ‘The Zoomorphic MiRo Robot’s Affective Expression Design and Perceived Appearance’ explores the design of social robots and its effectiveness.
We interviewed Professor Moojan Ghafurian from the Department of Electrical and Computer Engineering to learn more about her experience of using MiRo for her research.
Images were taken from the University of Waterloo.
Moojan, can you tell us about you and your research team?
Professor Moojan Ghafurian, the University of Waterloo
We are Moojan Ghafurian, Gabriella Lakatos, and Kerstin Dautenhahn. Moojan Ghafurian is a Research Assistant Professor in the Social and Intelligent Robotics Research Laboratory (SIRRL) at the University of Waterloo. Gabriella Lakatos is a Lecturer at the School of Engineering and Computer Science and a member of the Adaptive Systems Research Group at the University of Hertfordshire, in the UK. Kerstin Dautenhahn is the Canada 150 Research Chair in Intelligent Robotics at the University of Waterloo and the co-founder of the Social and Intelligent Robotics Research Laboratory (SIRRL).
This article used our team’s expertise in human-robot interaction, human-computer interaction, social robotics, and dog behaviour.
Can you tell us about your project and what you were trying to learn?
We work on increasing social capabilities of social robots, which can be used in multiple domains such as healthcare and education. One of these capabilities is the ability to show and communicate through emotions. Designing emotional displays of social robots is challenging due to multiple limitations that each robot may have, and it gets more challenging when implemented on an animal-like robot, as multiple factors can affect people’s understanding of animals’ emotions. Our study designed and evaluated 11 affective expressions, including emotions and moods, for the MiRo robot, which, to some extent, can be applied to other zoomorphic robots.
Why did you use MiRo for your project?
We used MiRo as we were looking for an expressive zoomorphic robot that may not necessarily look like any specific animal. This becomes especially important in situations where those interacting with the robot may have a fear of a specific animal.
How did you go about delivering your project?
We designed and implemented the affective expressions of MiRo, which were inspired by the behaviour of dogs and other animals, as well as humans and cartoon characters. We then evaluated these affective expressions in user studies. We also studied how MiRo’s appearance is perceived in general.
“[MiRo] is a great platform that we plan to use in the future studies.”
Did you experience any challenges? If so, what were they?
Similar to the other robots, MiRo has limitations that made it challenging to implement some of the affective expressions on it. One challenge was that the ears cannot be bent, which is an important factor for showing some emotions in dogs and other animals. The other challenge was that changes in facial expressions of MiRo are limited to manipulating the eyelids. Therefore, we added multiple different body motions to help convey some emotions.
What were the results?
We provided design guidelines for 11 affective expressions that can be applied, to some extent, to other zoomorphic robots. The user studies supported that the designed expressions, especially six of the affective expressions, could be recognized well. We also found that the general ability to recognize humans’ and animals’ emotions affected participants’ recognition of MiRo’s emotions. Further, MiRo was perceived to be gender neutral, and, while it was often seen as a dog or a rabbit, it was also perceived to be a variety of other animals, such as a mouse, a cat, a cow, and a donkey.
What did you like about using MiRo?
We found MiRo’s range of motions to be a great advantage in designing affective expressions that are shown through head and body movement, such as its expressive abilities through tail wagging, blinking, rotating ears, and head movements. Overall, it is a great platform that we plan to use in the future studies.
What are the next steps or further developments from this project?
We aim to use MiRo and its affective expressions in different domains, such as for supporting older adults. We also investigated the recognition of these affective expressions in conditions with limited visibility and showed that recognition of these affective expressions could be quite robust in conditions.
