NEWS
October 28, 2024
The other day, on a call with a colleague, he said, ”Wait a second, I have to turn off Tøffe, he’s making too much noise.” I immediately thought Tøffe was the family pet, but thought it was a little bit weird that he had to be ”turned off.”
It turned out that Tøffe was the family’s brand-new robot vacuum cleaner.
Not only had they named it, they also referred to it as a he. They had completely and pretty much instantly begun to think of it as lifelike. They had socio-morphed it, as recent Ph.D. graduate Rebekka Soma would say.
It turns out that this reaction isn’t uncommon.
Have you ever watched a robot move and felt a flicker of recognition, a sense that it was somehow alive? Even if a robot looks nothing like a human, its ability to move can be enough to make us perceive it as lifelike.
In this edition of Robots & Humans, we talked to Rebekka Soma from the University of Oslo, about her work on robots and why movement is such a powerful cue for our tendency to socio-morph robots and see them as lifelike.
As Rebekka puts it, ”Movement is fundamental to our understanding of life. From the moment we are born, we experience the world through movement. We learn about our bodies and our surroundings by interacting with them physically.”
As philosopher Maxine Sheets-Johnstone argues, ”We come straightaway moving into the world.” This primal connection between movement and being means that we instinctively associate movement with life.
When we see a robot moving, our brains automatically try to make sense of its actions. We look for patterns and try to understand its goals. This process of interpretation is similar to how we understand the movements of other humans and animals. We instinctively see movement as intentional, even if we know that the robot is simply following a pre-programmed set of instructions.
We feel like Tøffe, the robot vacuum cleaner, is determined in his mission to scour every corner of the room, removing all dust and debris before returning to base. But in fact, Tøffe doesn’t really know anything about cleaning the floor. He's just doing what he’s been programmed to do – move until he encounters an obstacle and then move away from it.
Our tendency to perceive robots as lifelike is heightened when their movements are expressive. Robots that move in a fluid and dynamic way are more likely to elicit feelings of animacy than robots that move in a jerky or mechanical way. This is because expressive movement suggests agency - the ability to act independently and make choices.
Researchers in human-robot interaction are exploring how to design robot movements that are both functional and expressive. They are drawing inspiration from fields like animation and dance to create robots that move in a way that is both natural and engaging. For example, using animation principles like "slow in and slow out" can make a robot's movements appear smoother and more lifelike.
However, even robots with limited movement capabilities can seem lifelike. As the sources note, a robot that simply pauses or rotates can evoke a range of interpretations from observers. Like the little delivery robot in the first installment in this series, we may see these movements as signs of hesitation, confusion, or even contemplation.
This highlights the power of movement - the idea that every movement, even a non-movement, can communicate something to an observer.
Ultimately, our perception of robots is shaped by our own embodiment. We understand the world through our own bodies and experiences. When we see a robot moving, we project our own understanding of movement onto it. This can lead us to see robots as more lifelike than they actually are.
This has implications for a wide range of issues, not the least of which is safety. Sonair is using ADAR technology to develop sensors that help robots ”see” better and thus become safer, which has a positive impact on safety.
As robots become more prevalent in our lives, it's important to be aware of how our perceptions of them are shaped by our own biology. By understanding the power of movement, we can better design and interact with robots in a way that is both beneficial and meaningful.
We’ll have more of our conversation with Rebekka in the next Robots & Humans, as there’s much more to explore in her very interesting work.
**
Rebekka Soma is a recent Ph.D. graduate from The Faculty of Mathematics and Natural Sciences at the University of Oslo. Her Ph.D. thesis is entitled "Does a Robot Lawnmower Know Anything about Grass?" We spoke to her about her research for our series Robots & Humans.