Design with a pet dog analogy for intimate interaction of smart products

Abstract

As smart products and systems are beginning to penetrate deeply into people’s daily lives, researchers have concerns about whether highly intelligent products can degrade user experiences. A system’s complexity can decrease both the usability and user acceptance of a new technology, which results in an emotionally poor experience and user dissatisfaction. Due to the recognition of the importance of this issue, various theories on models for designing future smart products have been introduced. One notable implication of this is that recent studies are moving toward bringing animistic connotations to smart product design. However, there is sparse research on how to employ animistic attributes to smart products in practical design processes. More research is needed to provide a sufficient knowledge framework, and it is necessary to have the practical tools to leverage an animistic approach for smart product design. In line with previous visions of animistic connotations, I paid attention to the analogy of pet dogs as an animistic design exemplar for smart products capable of intimate interactions. I define the term “intimate interaction” as the concept of interactions that actively form relationships with users by inducing positive and loving feelings as well as emotional experiences through personal artifacts. In this dissertation, the pet dog analogy refers to design attributes that assume that an intelligent device acts and interacts like a pet dog. The primary aim of this research is to investigate a systematic method and tools for the intimate interaction of smart products with a pet dog analogy. In this study, I explored the design method based on its application to a smart car interface. To glimpse the potential of using a pet dog analogy for smart product design, I carried research through design (Zimmerman, 2007). A pet-like smart PC peripheral named CAMY was developed by bestowing affective pet dog characteristics that were based on secondary research. The results of this empirical study revealed the positive impact of pet-like smart product interfaces on intimate interactions and emotional user experiences. The CAMY study led me to explore the method of designing with a pet dog analogy through expressive and behavioral design properties. To investigate the expressive design properties, I selected 27 representative interactive objects that provided lifelike impressions and conducted a workshop to analyze the motives that triggered the lifelike impressions in each case with 15 design researchers. By using the triading method (Hanington, 2012), I derived the key characteristics of the product representations that prompted the primitive perception of living things. These characteristics were articulated as pet dog expressive attributes (PEAs). PEAs are composed of basic, dynamic, visual, tactile, and acoustic attributes. Regarding the behavioral design properties, two studies were conducted: an exploratory interview and an online survey. The exploratory interview was conducted with 12 dog owners to elicit the key characteristics of pet dog behavior that could be applied to the interactions of a smart car interface. These characteristics were articulated as pet dog behavioral traits (PBTs). PBTs are composed of self-expression, empathy, faithfulness, and innocence. This was followed by an online survey to investigate whether each PBT was more or less appropriate for various use situations of a smart car. Based on the expressive attributes and the behavioral traits, I proposed a design framework for using a pet dog analogy. The design framework displays the two types of design properties (i.e., PEA, PBT) and their design targets (i.e., design elements, design contexts) where the design properties are applied. This framework also shows the relationship between design properties and design targets. To augment the applicability of the framework in design practice, I developed a practical design tool named Petmorph. It consists of a web-based application that easily explores the relationship between design properties and design targets and a guidance document that helps users use the application in the design process. To verify the effectiveness of Petmorph, I conducted two studies: a design workshop and a case evaluation study. 14 design students participated in the workshop, and the design outcomes were evaluated by 11 design experts with regard to the possibility of intimate interactions. In the design workshop, Petmorph was generally used as a design option or strategy suggested by a trustworthy design guidance system. The tool was useful in generating concept ideas for intimate interactions in three ways: 1) in facilitating analytical design thinking, 2) in inducing ideas that involved more sensory and emotional feedback, and 3) and in making the design process efficient and confident. To assess the practical impact of a dog-like car interface case, a dog-like affective driving assistant (DADA) was implemented, and I evaluated its use experience through a user study with 13 participants. The results empirically indicated that a design case from Petmorph could provide intimate interactions. I also discussed some issues that can be raised when using Petmorph in design practice. First, I clarified the versatility of Petmorph and provided knowledge for its general use. Second, I provided the following suggestions to be considered when using it: 1) utilize the ambiguity of the representation, 2) consider the impact of the emotional burdens, 3) use active and dynamic representations carefully, and 4) consider a dog’s level of intelligence and ability. In addition, I articulated the impact of this design method on intimate interactions by focusing on three aspects of unique features: 1) intuitive and compelling but abstract, 2) emotionally appealing, and 3) dwelling between smart and dumb. Based on these, the potential of this method to be a pertinent model for smart product design and the value of its use were presented. I also discussed other issues raised in this dissertation. Though this dissertation focused mainly on its application to a smart car interface, this study’s implications and lessons can be extended to the design of general products. This paper will mainly contribute to the field of HCI and design in three aspects: 1) in proposing the possibility of design with a pet analogy to enhance the user experience of smart products, 2) in suggesting a practical tool for generating smart product concepts for intimate interactions, and 3) in providing implications for applying Petmorph in design practice. This research will inspire designers and researchers who are interested in applying a pet analogy to future products and systems by extending the boundaries of its applications and the necessary methodologies.