1 SMELL SPACE: Mapping out the Olfactory Design Space for 2 Novel Interactions 3 4 SCHI Lab, School of Engineering and Informatics, University of Sussex, UK 5 EMANUELA MAGGIONI, 6 ROBERT COBDEN, SCHI Lab, School of Engineering and Informatics, University of Sussex, UK 7 DMITRIJS DMITRENKO, SCHI Lab, School of Engineering and Informatics, University of Sussex, UK 8 KASPER HORNBÆK, Department of Computer Science, University of Copenhagen, Denmark 9 MARIANNA OBRIST, SCHI Lab, School of Engineering and Informatics, University of Sussex, UK 10 The human sense of smell is powerful. However, the way we use smell as interaction modality in HCI is 11 limited. We lack a common reference point to guide designers’ choices when using smell. Here, we map out 12 an olfactory design space to provide designers with such guidance. We identified four key design features: (i) 13 chemical, (ii) emotional, (iii) spatial, and (iv) temporal. Each feature defines a building block for smell-based 14 interaction design and is grounded in a review of the relevant scientific literature. We then demonstrate 15 the design opportunities in three application cases. Each application (i.e. one desktop, two virtual reality 16 implementations) highlights the design choices alongside the implementation and evaluation possibilities in 17 using smell. We conclude by discussing how identifying those design features facilitates a healthy growth of 18 this research domain and contributes to an intermediate-level knowledge space. Finally, we discuss further 19 challenges the HCI community needs to tackle. 20 CCS Concepts: • Human-centered computing → Interaction techniques; Empirical studies in HCI. 21 Additional Key Words and Phrases: Smell; Olfactory Design Space; Chemical Sense; Scent-based Interaction 22 Design; Odour Interfaces; Smell-based Applications; Novel Interactions 23 ACM Reference Format: 24 Emanuela Maggioni, Robert Cobden, Dmitrijs Dmitrenko, Kasper Hornbæk, and Marianna Obrist. 2019. SMELL 25 SPACE: Mapping out the Olfactory Design Space for Novel Interactions. ACM Trans. Comput.-Hum. Interact. 26 37, 4, Article 111 (August 2019), 26 pages. https://doi.org/10.1145/1122445.1122456 27 28 1 INTRODUCTION 29 Consider a future where your nose has become as important as your eyes in reading this sentence. 30 Our sense of smell can help perceive information when our visual system is busy (e.g. visual 31 overload) or unusable (e.g. in darkness). Similar to visual and also auditory stimuli [110], scents 32 exist spatially. The presence of a scent source (i.e. a scent stimulus) and its diffusion can be located 33 in space (i.e. scented air volume) even if out of sight [103]. Moreover, it has been shown that scents 34 can convey meaning and complement visual information processing [101] and decision making 35 36 Authors’ addresses: Emanuela Maggioni, SCHI Lab, School of Engineering and Informatics, University of Sussex, UK, Chichester 1, BN1 9QJ Brighton, UK, [email protected]; Robert Cobden, SCHI Lab, School of Engineering and 37 Informatics, University of Sussex, UK, Chichester 1, BN1 9QJ Brighton, UK, [email protected]; Dmitrijs Dmitrenko, SCHI 38 Lab, School of Engineering and Informatics, University of Sussex, UK, Chichester 1, BN1 9QJ Brighton, UK, D.Dmitrenko@ 39 sussex.ac.uk; Kasper Hornbæk, Department of Computer Science, University of Copenhagen, Denmark, Nørregade 10, 1165 40 København, Denmark, [email protected]; Marianna Obrist, SCHI Lab, School of Engineering and Informatics, University of 41 Sussex, UK, Chichester 1, BN1 9QJ Brighton, UK, [email protected]. 42 Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee 43 provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and 111 44 the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires 45 prior specific permission and/or a fee. Request permissions from [email protected]. 46 © 2019 Association for Computing Machinery. 47 1073-0516/2019/8-ART111 $15.00 48 https://doi.org/10.1145/1122445.1122456 49 ACM Trans. Comput.-Hum. Interact., Vol. 37, No. 4, Article 111. Publication date: August 2019. 111:2 Maggioni, et al. 50 (e.g. enabling cognitive shortcuts [20]). Above all, our sense of smell has a unique and robust link to 51 emotions, and may make experiences more memorable [21, 109]. Despite the sense of smell being 52 very powerful, its use within HCI is still very limited. 53 Within HCI, the use of scent for interaction and experience design is increasingly recognised (e.g. 54 [68, 74, 97, 104]). Most efforts are however directed towards the development of novel scent-delivery 55 devices (e.g. [1, 29, 32]) and the design of one-off application examples, such as the integration of 56 scent-based interactions into virtual and augmented reality (e.g. [104] and [93]), gaming [90], and 57 multimedia applications [43, 45, 89]. The most common motivation for designing with smell is to 58 create more immersive experiences, mainly referring to concepts such as the sense of presence, 59 immersion, and realism [3, 50, 64, 105]. In addition, we see attempts to study the effect of scent 60 stimuli on emotions (e.g. to reduce stress [2, 127]) and behaviour (e.g. reduce distraction, help 61 multi-tasking [52, 58, 82]). All those efforts demonstrate the desire of the HCI community toextend 62 interaction design beyond the audio-visual domain. Moreover, it underlines the opportunities that 63 our sense of smell provides application designers. However, to move the use of smell beyond one-off 64 interaction examples, we need to establish a common reference point that enables designers to 65 make informed decisions about the use of smell as interaction modality. 66 Based on a detailed literature review, drawing upon advances on our understanding of the 67 olfactory system in psychology, neuroscience, sensory science, and biology, we identified four key 68 design features for smell: chemical, emotional, spatial, and temporal. Those four features define the 69 building blocks for the olfactory design space we introduce in this article. We then discuss how 70 to navigate this design space by following a Design Space Analysis approach [81] that highlights 71 Questions, Options, and Concerns (Q-O-C) as key anchor points. This rationale-based approach 72 helped us to formulate specific questions linked to designing with smell and consider options 73 based on specific concerns we highlight for three application cases. The application designs and 74 evaluations include one desktop implementation (i.e. messaging system in a work context) and two 75 virtual reality (VR) implementations (i.e. time management game in VR, localisation task in VR). 76 In summary, the main contributions of our work are threefold: First, we identify four key design 77 features (i.e. chemical, emotional, spatial, temporal) that help map out the olfactory design space 78 for HCI and thus provide designers with a common reference point when designing with smell. 79 Second, we demonstrate the relevance of those features in the design process that includes the 80 implementation and evaluation of three application cases (i.e. one desktop and two VR applications). 81 Finally, we discuss how our theoretical and empirical exploration of smell as novel interaction 82 modality enriches the audio-visual design space in HCI and adds new intermediate-level knowledge 83 (i.e. bridging theory with practice) to the design of future smell-based interactive systems. We 84 conclude by accounting for design trade-offs and the need for further research to move smell 85 beyond its infancy state of today. 86 87 2 OLFACTORY DESIGN SPACE 88 Despite the complexity of the sense of smell (i.e. detection of chemicals in our environment), it is 89 astonishing how well humans can react to and act on scent stimuli. We intuitively interact with the 90 environment and the molecules in the air, making decisions beyond pure hedonic discrimination 91 (i.e. pleasant/unpleasant scents). Scent can increase the saliency of an object and can facilitate its 92 recognition and categorization [20]. For example, smelling coffee causes the mental representation 93 of coffee to be activated. This can lead to a desire for coffee, or, with the implicit association between 94 coffee and breaks, motivate us to have a break. At the same time, information delivered byscents 95 can go beyond simple association and activate an instinctual behavioural reaction. For instance, 96 recognising danger in the scent of gas (i.e. the smell of ‘sulphur’) triggers our survival instinct [67] 97 and promotes actions to protect ourselves (e.g. open a window, leave the room). 98 ACM Trans. Comput.-Hum. Interact., Vol. 37, No. 4, Article 111. Publication date: August 2019. SMELL SPACE: Mapping out the Olfactory Design Space for Novel Interactions 111:3 99 100 101 102 103 104 105 106 107 108 109 Fig. 1. Mapping out the Olfactory Design Space based on four key features (i.e. chemical, emotional, spatial, 110 temporal) extracted from a systematic literature review on smell as future interaction modality in HCI. 111 112 113 The olfactory system has an eminent importance in classifying scents into the food or non-food 114 category [7]. Apart from distinguishing edible sources, the availability of food (e.g. scent of freshly 115 baked bread in western cultures), also triggers the human survival instinct and attracts us towards 116 the source of the scent (e.g. salivation [7, 17, 73]). This instinctive behaviour does not necessary 117 mean that we are at the edge of starvation, but it is a biological behaviour, automatic reaction to 118 scents.