What Can We Learn About Dog and Animal Behavior with Robots?

An issue of Animal Behavior and Cognition considers this forward-looking topic.

Posted Nov 11, 2018

"Animal-Robot Interaction (ARI) is not a novel approach in the study of animal behavior and cognition, but despite the methodological advances it provides, it is not as widespread as one may expect. The use of interactive, autonomous robots can give benefits which go well beyond the opportunities offered by early stimulus models. Such autonomous  robots can help collecting data about animal behavior that is not possible achieve with any other methods. We hope that this collection of reviews and research studies may provide a new push to move this field forward." ("Methodological Challenges of the Use of Robots in Ethological Research")

I recently learned of a fascinating special issue of the journal Animal Behavior and Cognition edited by Adam Miklósi and Judit Abdai devoted to the topic of robots and nonhuman animals (animals). Among the animals considered are fishes, rats, and dogs (please also see), and all of the essays are available online for free. This is the first edited volume devoted to this subject, often called ethorobotics (for a review please see "Ethorobotics: A New Approach to Human-Robot Relationship").

Of course, ethorobotics is not a substitute for humans carefully observing and studying other animals, but it's clear we can learn a lot from the use of robots in our attempts to learn more about a wide variety of animals. From time to time I've thought about different aspects of robot-animal interactions and what we can learn by using robots to study various aspects of animal behavior, but I've always been distracted by other things and came to realize that I really didn't know enough about some of the nitty-gritty methodological details. So, when I began reading the essays in this issue, I figured that I might as well go to the experts and gladly, both Adam and Judit were willing to answer a few questions. Here's what they had to say about this very interesting and forward-looking topic. 

Both of you recently edited a special volume of Animal Behaviour and Cognition on the general topic of robots and animals. Why do think this topic is important and how did you become interested in it?

Adam: Quite independently, a few years ago a couple of researchers had started to use robots to study animal behavior. That was the beginning of a new interdisciplinary research area of animal-robot interaction (ARI). Their goals were rather different. Some were interested in using robots as shoal mates in social fish to find out how individual decisions influence group behavior. Others were using robots to study the information transfer of food location in honey bees. Our idea was to deploy simple robots to find out whether after specific interactive experience dogs may react towards them as a potential social partner.

How did you select the essays that were included in this special issue as representative of what can be done using animal-robot interactions? 

Judit: As the small number of papers in the special issue may give a hint, finding manuscripts was not an easy task. As far as we know this is the first edited volume on this topic. This is a very new field, and ethologists are somewhat reluctant to start research including robots of any kind. Luckily, however, we ended up with a nice variety of topics. This may strengthen the view of our readers that animal-robot interaction – as a method – has the potential to be applied widely across behavior and cognitive studies. Unfortunately, regarding the species, the variety is smaller, but there are several studies applying animal-robot interactions in insects, birds, and mammals.

Adam Miklósi
Dog doing a play bow to a car
Source: Adam Miklósi, used with permission

How does the use of robots open the door for studies that have not been conducted before -- how does it advance the study of animal behavior and animal well-being? Can you relate this to the use of robots to study dogs, because so many people will be interested in dog-robot interactions (DRIs)?

Adam: We believe that in the longer run, the use of interactive robots will change the research methods in the behavioral sciences, including ethology. Consider, for example, that it is rather difficult to study social interaction between individuals because the researcher has little control over what is happening. If these interactive robots can play a role of a social partner, then their action and reaction can be programmed by the scientists in advance.

From a practical point of view, such interactive robots can be also deployed as social partners for dogs and cats. Importantly, they should not replace the owners and these family companions will always require social interaction with humans, but these devices can take the role of a playmate.

What are some of the shortcomings, challenges, and benefits of using robots to study animal behavior?

Judit: ARI has several difficulties, but all novel methodologies present some challenges. If one prefers to use a commercially available robot, the job is easier because one could expect that it will work (however, such expectations sometimes do not meet reality). Building a new one is also a possibility but this takes usually more time than expected. And unfortunately, just like animals, even robots have the habit of not behaving in a way we expect them to do. However, unlike in the case of animals, we have a chance to fix the problem. But this is worth the effort because there are almost unlimited options to use interactive robots. We can take away or add new features at any time, and the robot will do what we ask and when we ask it.

What are some of the practical implications of using robots for people who choose to share their homes with a companion animal? How can people use robots and the results of these studies to help them understand what their nonhuman companions need in order to give them the best lives possible?

Adam: Interactive robots have several advantages for being included in human homes but people also have to learn how to use them. Importantly, we are a good 5-10 years away from this situation becoming reality but this also gives us time to find out a lot about ARI. In principle, interactive robots can be as helpful for people as for their companions. So one of the main challenges for these robots is to be able to navigate in the complex social network of a human family. These robots should be never seen as being a replacement for humans, dogs or cats. Such interactive robots should be regarded as representatives of a new species that has the capacity for social interaction with others. In specific situations, such as dogs showing separation-related behavior, such interactive robots may have the potential to eliminate the problem.

What do you see as the future of studying animal-robot interactions? Is this methodology limited to certain species?

Judit: We believe that interactive robots can be used widely, both from the viewpoint of an investigated question and the species. As you can see in the current literature, robots have already been used to study the behavior of bees, cockroaches, fish, lizards, birds, squirrels, rats and dogs, among others. We think that finding the appropriate robot can be sometimes difficult (for some studies they have to look exactly like the studied species, while in others the displayed behavior is more important), but one can learn a lot even during these initial steps. We truly believe that in the future ARI will be a widespread method that may facilitate not only studies within species, but allow a better approach to comparative investigations as well. There is already a newly established field, called ethorobotics that will harbor both theoretical and methodological approaches.

What are some of your current and future projects in this area?

Judit: Currently, we use robots to study the important features in animacy perception, and the behaviors and cognitive capacities that help dogs to accept an agent as a social partner. We have further plans on investigating long-term memory, cognition and social behavior in dogs, substituting humans and conspecifics with specific robots as social partners. And soon cats will also be introduced to our robots.

Thank you very much for this very interesting interview. I hope more and more researchers will consider using robotics to learn more about the fascinating animals with whom we share our magnificent planet. Of course, ethorobotics is not a substitute for humans carefully observing and studying other animals, but it's clear we can learn a lot from the use of robots in our attempts to learn more about a wide variety of animals. 

More Posts