Towards Intelligent Playful Environments for Animals based on Natural User Interfaces
With the aim of improving animals’ wellbeing, my PhD research proposes the design and development of technological playful environments able to adapt and react to the animals’ playful interactions, allowing them to play either by themselves or with a human and providing more engaging and dynamic playful activities that can be adapted over time.
Inspired by the emergence of Natural User Interfaces for human users, my thesis explores the design of Natural User Interfaces for animals that support the development of interactive spaces for them, in which humans could also participate as players and not just as providers of the activity. These NUIs for animals (and human) participants would be based on the study of their body movements, gestures or actions during the playful activity. For this, this thesis resorts to non-wearable tracking systems as they can help to detect animals’ spontaneous behavior without the limitations or changes in behavior that the use of wearables might cause. In addition, this tracking approach would allow us to extract information not only from the animal player but also from the rest of the environment, i.e. interactive devices or human participants and their interactions with the animals.
- Patricia Pons, Javier Jaén. Enabling Awareness in Playful Environments for Animals using Body Tracking. AWARE Workshop on Awareness Interfaces & Interactions – IUI 2017.
- Patricia Pons, Javier Jaén. Enabling Embodied Interactions in Playful Environments for Animals. Proceedings of The Third International Congress on Animal Computer Interaction – Doctoral Consortium, 2016.
Depth-based cat tracking system
If we want to develop environments whose interactive devices adapt to the animals’ behaviors and preferences, we need a way of capturing the animals’ movements unobtrusively (without using a harness nor a wearable device, allowing the animal to move and interact freely and in a natural way). For this purpose, we have developed the first depth-based tracking system capable of detecting a cat’s position, body posture and orientation using computer vision and machine learning techniques. This tracking system would allow us to create engaging and realistic games using technological artifacts which adapt to the detected cat’s postures and field of view.
- Patricia Pons, Javier Jaén, Alejandro Catalá. Assessing machine learning classifiers for the detection of animals’ behavior using depth-based tracking. Expert Systems with Applications (in press).
- Patricia Pons, Javier Jaén, Alejandro Catalá. Developing a depth-based tracking systems for interactive playful environments with animals. 12th International Conference on Advances in Computer Entertainment Technologies – Second International Congress on Animal Human Computer Interaction, 2015.
Tangible User Interfaces for Animal Enrichment
Study and recognition of cats’ playful preferences and behavior
In order to study the most suitable mechanisms to create engaging playful experiences for cats, we conducted an observational study to evaluate the interest of several cats of different ages, personalities and living conditions, when facing different kinds of technologically-based stimuli and interaction modalities. With the assistance of experts in cat behavior and cat therapists, we observed different behaviors and preferences on the cats when playing with the digital stimuli. We analyzed the factors and dimensions that could have influenced the interactions, such as age, environment, past experiences and personality traits.
- Patricia Pons, Javier Jaén, Alejandro Catalá. Towards Future Interactive Intelligent Systems for Animals: Study and Recognition of Embodied Interactions. Proceedings of the 22nd International Conference on Intelligent User Interfaces (IUI 2017), pp. 389-400, 2017.
- Patricia Pons, Javier Jaén. Towards the Creation of Interspecies Digital Games: An Observational Study on Cats’ Interest in Interactive Technologies. Proceedings of the 2016 CHI Conference Extended Abstracts on Human Factors in Computing Systems, pp. 1737-1743, 2016.
Tangible interactions for zoo animals
During 2016 I did a research stay at the Microsoft Research Centre for SocialNUI at the University of Melbourne. In collaboration with Zoos Victoria, we focused on the design and development of a depth-based tracking system capable of detecting the orangutans’ movements of everyday non-technological objects. These interactions were then digitally augmented with technological stimuli, such as sounds, to create varied and adaptable enrichment activities, e.g. produce different sounds based on the orangutans’ movement of everyday non-technological objects detected by the system. The evaluation of this system would potentially allow to study which kind of auditory stimuli are more engaging for them and which mappings would be best to mediate the interactions. The prototype system is yet to be evaluated.
- Patricia Pons, Marcus Carter, Javier Jaén. Sound to your Objects: A Novel Design Approach to Evaluate Orangutans’ Interest in Sound-based Stimuli. Proceedings of The Third International Congress on Animal Computer Interaction, 2016.
Interspecies Playful Interactive Environments: Design and Development of Human-Animal Playful Experiences
Human participation in playful environments for animals could help to improve not only the wellbeing of animals and humans, but also their interrelationships and bonds by creating a shared space for their amusement. In this regard, it is essential that future generations develop empathy and create strong bonds with the animals in our ecosystems to work towards a more inclusive world for all species. Our last studies have explored how humans, and especially children, envision their participation in such interactive environments, how NUIs could support these interactions, and how interspecies playful interactive environments can help to improve the development of intelligent and adaptive systems for animals while helping children’s wellbeing and relationships with them.
We have designed, developed and evaluated one of the first animal-centered interspecies remote playful system for animals and humans, that builds upon the insights obtained from the different studies in my PhD research. This system allows hospitalized children to play remotely with an animal located in a daycare facility. Children use a mobile application to control a tangible robot in order to play remotely with a dog located in a canine daycare facility. An autonomous interactive system with basic intelligent features for animal play has also been designed and evaluated, allowing animals to play by themselves with the robot and the system. This work compares the two different approaches in terms of their suitability, usability and human/animal engagement, identifying behavioral patterns and future promising improvements.
For a complete list of publications of all these projects, click here.