A portrait of a happy leonberger with beautiful brown eyes
One time when I took my dog for a walk, someone was playing with a radio-controlled car on the road. This was Very Exciting (actually most things in life are Very Exciting!). I asked him to heel as we walked past it, but he would have preferred to investigate the car. How do dogs interact with things like this, and does it change if the object moves in an apparently social way?

Robots have been used in studies with various species as a way of experimentally testing the rules of communication. Perhaps the most interesting for dog lovers is Leaver and Reimchen’s (2008) study in which off-leash dogs interacted with a dog-like robot with one of four different possible tails: short and still, short and wagging, long and still, or long and wagging. With the long tail, dogs were more likely to approach if it was wagging rather than still, but there was no difference between the two short tail conditions.

A new study by lead author Anna Gergeley et al in Hungary tests whether the actions of a mechanical object affect how a dog responds. They used a radio-controlled car as an Unidentified Moving Object. In one condition, the car (operated by an experimenter) always followed a fixed route in the lab. They called this the Mechanical UMO. In another condition, the experimenter made the car act as if it was social, by making it move in response to the dog looking at it (Social UMO). The car also had eyes drawn on the windscreen. In a third condition, a human played the role of a Mechanical Human by wearing sunglasses (so as not to make eye contact) and following exactly the route of the fixed UMO in a mechanical way.

The photograph below shows a) Mechanical UMO; b) Social UMO; c) Mechanical Human.

Forty-seven pet dogs took part in the study (a further three were recruited but had to be excluded due to anxiety). They were divided into three groups, and each dog took part in just one of the conditions. 

Source: PLoS One

After an initial warm-up, there was a set of trials which began with an experimenter placing food into one of three bowls in the room. The dog stayed by the owner at one end of the room. The Unidentified Moving Object or the Mechanical Human set off to move about the room, taking the food from the bowl and placing it in a cage where it was out of the dog’s reach.  Then the dog was released to explore for a fixed period of time, and then called back to the owner. The UMO or Mechanical Human began to move again, this time collecting the food from the cage and delivering it to the dog. An ingenious set-up with magnets allowed the UMO to move the food around.
 
The results showed that more dogs looked at the UMO in the Social condition than when it was Mechanical, or the Mechanical Human. This suggests that the behaviour of the item affects how dogs respond to it. The dogs also spent more time looking at the Mechanical UMO rather than the Mechanical Human, possibly because it was a novel item.

As in Leaver and Reimchen’s study, this shows that it is possible to test how dogs respond to apparently-communicative behaviour by using a mechanical object. Communication in the real world is messy, so experimental set-ups like this are sometimes very useful.

This study makes me wonder about the possibilities for hi-tech interactive dog toys. The electronics would have to be in a chew-proof housing, of course!

How does your dog (or cat) respond to a remote-controlled car?

References
Gergely, A., Petró, E., Topál, J., & Miklósi, A. (2013). What are you or who are you? The emergence of social interaction between dog and an Unidentified Moving Object (UMO) PLoS ONE 
Leaver, S.D.A., & Reimchen, T.E. (2008). Behavioural response of Canis familiaris to different tail lengths of a remotely-controlled, life-size dog replica Behaviour, 145 (3), 377-390


Leave a Reply