Expected user experience of mobile augmented reality services: a user study in the context of shopping centres (original) (raw)

Expected user experience of mobile augmented reality services: a user study in the context of shopping centres

Thomas Olsson ⋅\cdot Else Lagerstam ⋅\cdot Tuula Kärkkäinen ⋅\cdot Kaisa Väänänen-Vainio-Mattila

Abstract

The technical enablers for mobile augmented reality (MAR) are becoming robust enough to allow the development of MAR services that are truly valuable for consumers. Such services would provide a novel interface to the ubiquitous digital information in the physical world, hence serving in great variety of contexts and everyday human activities. To ensure the acceptance and success of future MAR services, their development should be based on knowledge about potential end users’ expectations and requirements. We conducted 16 semi-structured interview sessions with 28 participants in shopping centres, which can be considered as a fruitful context for MAR services. We aimed to elicit new knowledge about (1) the characteristics of the expected user experience and (2) central user requirements related to MAR in such a context. From a pragmatic viewpoint, the participants expected MAR services to catalyse their sense of efficiency, empower them with novel context-sensitive and proactive functionalities and raise their awareness of the information related to their surroundings with an intuitive interface. Emotionally, MAR services were expected to offer stimulating and pleasant experiences, such as playfulness, inspiration, liveliness, collectivity and surprise. The user experience categories and user requirements that were identified can serve as targets for the design of user experience of future MAR services.

Keywords Augmented reality ⋅\cdot Mobile services ⋅\cdot Ubiquitous computing ⋅\cdot User experience ⋅\cdot User expectations ⋅\cdot User requirements

[1]

1 Introduction

Integrating information processing in everyday objects and making the physical world a user interface for digital information have been popular targets of research during the last two decades. The paradigm shift from desktopbased interaction towards mobile and ubiquitous computing, happening “anytime, anywhere” [46, 54], is gradually becoming a reality in people’s lives. Our personal devices are being turned into external eyes and ears for sensing embedded information in the surrounding environment. The strengths of both the real and digital worlds can be integrated in a single interface, which enables novel applications and services to be developed. Consequently, this creates both possibilities and challenges for the design of pleasurable user experience.

Augmented reality (AR), as an enabler of ubiquitous computing and location-based services, can be considered as one key approach in catalysing the diffusion of technological approaches such as ambient intelligence [40], location awareness and the internet of things. The ongoing rise of AR has become evident with a large body of publications describing MAR demonstrators, as well as the recent introductions of publicly available mobile AR applications, such as Layar, 1{ }^{1} Wikitude 2{ }^{2} and Junaio 3{ }^{3} [37]. Despite the recent advent of various AR demonstrators, user experience and users’ expectations of MAR services have received little consideration in the research and, because of their salience in the design of successful services, deserve more attention.

[2]

1. T. Olsson ( ⊠)⋅\boxtimes) \cdot E. Lagerstam ⋅\cdot T. Kärkkäinen
   K. Väänänen-Vainio-Mattila
   Unit of Human-Centered Technology, Tampere University of Technology, Korkeakoulunkatu 6, 33720 Tampere, Finland
   e-mail: thomas.olsson@tut.fi ↩︎
2. 1{ }^{1} www.layar.com.
   2{ }^{2} www.wikitude.org.
   3{ }^{3} www.junaio.com. ↩︎

1.1 Framing the concept of mobile augmented reality

The general approach of AR is to combine real and com-puter-generated digital information into the user’s view of the physical real world in such a way that they appear as one environment [22, 52]. This new reality of mixed information can be interacted within real time, enabling people to take advantage of both their own senses and skills and the power of networked computing while naturally interacting in the everyday physical world [30].

Furthermore, the concept of AR relates to a broader concept of mixed reality (MR). Mixed reality refers to the integration and merging of the real and virtual worlds where physical and virtual objects complement and interact with each other [32]. AR is often understood as part of the mixed reality continuum (Fig. 1), focusing on augmenting the real world with digital information, instead of bringing real-world information in virtual worlds [2]. AR aims to supplement the real world, rather than creating an entirely artificial environment. To elaborate our scope in these broad research areas, this paper focuses especially on mobile augmented reality (MAR), that is, AR created and accessed with mobile devices in mobile contexts of use.

Mobile devices, such as mobile phones, digital cameras and navigators, have become a fruitful platform through which to apply mixed and augmented reality technologies. Because of the rapid development of sensor and communication technology, mobile devices are becoming increasingly “aware” of the information resources and affordances nearby, the user’s social network, device orientation, as well as other characteristics of the context [21, 38, 47].

By further increasing the sensing capabilities of technology, location awareness [25, 39] is another key component in any ubiquitous computing service. Hence, state-of-the-art mobile devices are starting to possess the required technical enablers for creating augmented reality services and interacting with the augmented information. Mobile location-aware AR services could work anywhere, visualising the digital layer of information related to the user’s real surroundings whenever desired. Hence, mobile AR can be extended to cover the rich diversity of scenarios and situations in mobile computing and thus provides an extensive design space for services.

A predominant way of interacting with AR information is the magic lens [2] with which the user may browse the
world and augmented information through a camera view. Head-mounted displays are often used to create a highly immersive and holistic experience of the AR view. Although AR is often regarded as very visual by nature, current mobile devices could also enrich the augmentation with auditory or haptic information. Consequently, the new realm of integrated digital and physical world information presented using AR holds the potential to revolutionise the way in which information is accessed and presented to people: “the world becomes the user interface” [22, 55].

Mobile augmented reality can potentially be applied in various practical day-to-day use cases, ranging, for example, from information acquisition and navigation to tourism, entertainment and social interaction. Examples of augmentable targets range from surrounding places, buildings, areas or other static targets to events, people, or other dynamic or abstract targets. The augmented information can be any information related to the above-mentioned areas: for example, basic description of the target, geographical information, temporally changing information like advertisements or socially constructed information, or entertainment. Overall, the gap between the real world and its digital counterparts or related metadata becomes narrower.

Considering the potential of AR to provide the user with relevant information and services [36], as well as the cognitively, emotionally and temporally changing situations in mobile contexts, the user’s experience with MAR services can become equally complex. To create successful MAR services, it is critical to understand what people expect of the experience in interaction with such services and the user requirements involved.

1.2 User experience and expectations

User experience (UX) takes a broad perspective on the user’s interaction with a product. UX moves beyond usability (i.e. effective, efficient and satisfactory interaction with the product [23]) towards more emotionally appealing relationships between the user and the product [7]. Several models and frameworks have been developed in the past decade to describe what UX is and how it is formed.

The definition of ISO states that UX is “a person’s perceptions and responses that result from the use and/or anticipated use of a product, system or service” [24].

Fig. 1 The reality-virtuality continuum: adapted from [32]

Hassenzahl and Tractinsky [18] have defined UX as “a consequence of a user’s internal state (predispositions, expectations, needs, motivation, mood, etc.), the characteristics of the designed system (e.g. complexity, purpose, usability, functionality) and the context (or the environment) within which the interaction occurs (e.g. organisational/social setting, meaningfulness of the activity, voluntariness of use)”. Overall, despite the research agenda of UX being still rather immature, some common key assumptions are widely accepted: the subjective nature, context-dependency and temporality [17, 27, 43].

In earlier research, the variety of experiences with interactive products has been much discussed. Desmet and Hekkert [6] distinguish components of product experiences as (1) aesthetic experience: delighting one or more of our sensory modalities, (2) experience of meaning: personal or symbolic significance of products and (3) emotional experience: emotions, such as joy and anger, elicited by the appraised relational meaning of the product. Hassenzahl [17] has described the diversity of experiences in terms of pragmatic (goal-oriented) and hedonic (action-oriented) motivations and product characteristics [19]. He has identified three main categories of product qualities that can produce positive hedonic user experiences. Stimulation is the ability of the product to provide new impressions, opportunities and insights to the user. Identification describes how users express themselves to others through using a product, and how people interact with others through the product. Evocation refers to the user’s ability to go back to positive memories with the help of the product.

Overall, pleasurable and rich user experience has gained ground as the foremost goal in the design of interactive systems. Hassenzahl et al. [18] point out that UX focuses on positive outcomes, such as positive emotions, of interaction. Desmet and Hekkert [6] state that positive emotions broaden people’s focus, facilitate discovery and increase well-being. Thus, the concept of UX moves the design focus from removing negative factors like usability or functionality problems to offering possibilities for positive experiences that exceed the user’s expectations.

When it comes to expectations, most UX definitions include the anticipated (that is, expected) use as an essential element of UX. In order to create successful and engaging products, UX should be taken into account in the early phases of design processes by systematically researching potential users’ expectations. In addition, the general user-centred design approach (UCD) [24] emphasises that new services in general should be based on true needs and expectations of the potential end-user group. User’s expectations reflect anticipated behaviour, thus having an influence on forming the actual user experience with the product [1]. Expectations influence the user’s perceptions of the product’s capabilities and quality, as
well as direct the user’s attention in interaction situations. If a product outperforms the expectations post-use, satisfaction [29] will result, whereas if it falls short of expectations, the user is likely to be dissatisfied [35].

From an experiential perspective, it is important to understand what kinds of experiences (i.e. the characteristics of experience) are seen desirable with a certain type of service or technology in a specific context or application area. In addition to theoretically understanding how the expectations influence the user experience and vice versa, it is important to actually gather the end users’ expectations [1]. This is essential with technologies or services that do not yet exist, that is, when there is a high risk of unsuccessful development investment.

1.3 Related research on user studies with AR applications

The research on AR has so far mostly focused on the development of enabling technologies-for example, various types of displays and other output devices (e.g. [32, 42]), and algorithms for identifying and tracking real-world objects (e.g. [2,53]) to integrate the real and the virtual. User research has been conducted mainly to evaluate the technical demonstrators, focusing on perception and cognition studies, user task performance or other usability-related aspects (e.g. [9,10,47])[9,10,47]), or to provide usability-oriented guidelines for design (e.g. [11, 13]). Overall, the whole approach of user-centredness has been underutilised in AR [13, 34, 49]. In the following discussion, we give a short overview of related applications and system prototypes that demonstrate mobile AR technologies and that have been evaluated with users.

A good example of an AR-based mobile application and a user study conducted on it is the ARCHEOGUIDE project presented by Gleue and Dähne [15]. ARCHEOGUIDE provides visual cultural heritage information augmented on a small mobile computer. They present a survey of mobile and wearable computing equipment and a description of the prototype development, as well as briefly discussing the experiences of using the system during the first trial phase. To mention a few other relevant studies, Sato et al. [45] present a novel mixed reality (MR) display system called MR-mirror. MR-mirror merges real visual information reflected on a real mirror and a virtual one displayed on an electronic monitor. Herbst et al. [20] present a mobile outdoor mixed reality game for exploring the history of a city in the spatial and the temporal dimension. The first one, Can You see Me Now? (CYSMN), was designed to be a fast paced game in which up to 20 online players were chased across a map of the city, by three runners who were moving through the actual city streets. Other AR solutions for the context of gaming are, for example, Touch-Space

[5], Penalty Kick [41], Impera Visco [41], Smart Memory [41] and Mah-Jongg [50].

A slightly more UX-oriented discussion on AR is presented by Barba et al. [3], who present a handheld AR system called [inbox] and discuss aspects like Physical-Virtual relationships, embodied interaction and increased awareness from an experiential point of view. Furthermore, Morrison et al. [33] present their findings from field trials of MapLens, which is an AR map using a magic lens over a paper map. The participants used the application to play a locationbased game in a city centre. Their main finding was that AR features facilitated place making by allowing referencing to the physical world. The strongest potential of AR maps was discussed to be their use as a collaborative tool.

Kuikkaniemi et al. [26] present the MAR Toolkit as an easy-to-use augmented reality toolset for building multiuser mobile phone games. They have demonstrated the MAR Toolkit by implementing and testing a game based on the classical board game “Scotland Yard”. Here, using graphical 2D-targets for providing location information for AR games was found to be a simple and robust alternative for more technology intensive GPS and cell-ID information.

Considering usability, a few studies have resulted in early guidelines for augmented and especially virtual reality. For example, Stanney [48] suggests taking into account user’s physical abilities (e.g. handedness), level of experience and technical aptitudes (e.g. orientation and spatial memory). Feiner [12] suggests supporting face-toface communication by presenting environment-related visual information within the user’s field of vision instead of requiring the user to look away, as well as to avoid interaction techniques that require a noticeable portion of the user’s attention. Wickens and Baker [56] suggest, for example, to support the user in orientation and localisation with accurate information and to strive for multi-modal and body-centred interaction. Nevertheless, many of these guidelines are still technology-centred and encompass very little of aspects related to user experience.

All in all, despite a large body of AR demonstrators and related research, the AR research community lacks understanding of what the user experience with mobile AR services could be, and what potential users expect of such services-especially regarding the emotional and hedonic elements of user experience. Such understanding, however, is important in order to be successful in developing wellaccepted services that provide the users with pleasurable and stimulating experiences.

2 Approach and methodology

Basing our approach on hermeneutics and a qualitative research stance, the overall objective of this study was to
understand potential users’ expectations of mobile augmented reality services. More specifically, we aimed at (1) understanding potential users’ views on what kind of user experience is expected from interacting with future mobile AR services and (2) gaining insight into user-originated requirements that would affect the user experience. Thus, we could identify categories of experiences that are worth targeting in design, and how to possibly reach them with designable service features. The topic was approached with a focus on MAR services (i.e. a holistic entity that encompasses functionalities, information and devices with which to access the features and information), as for end users it is often hard to distinguish between these aspects. The study took place during Spring 2009 when even the first publicly available MAR applications were still in their infancy and had reached only a small number of users. Therefore, we regarded it as more beneficial to focus on people’s expectations rather than on the early experiences with such immature applications.

The research objectives were approached with contextual interviews in authentic, specified, environments. Interviewing in context was expected to be more concrete than in laboratory environments and thus more fruitful in eliciting expectations and requirements, because it is based on “in-the-moment” experience and allows various realistic stimuli from an authentic context [4]. Examples of such stimuli in this case were the objects and places in the physical environment, the social and cultural context, and the activities and needs related to that context. We utilised semi-structured interviewing, as it is most useful in set-ups like this where broad issues are to be understood, but the range of respondents’ reactions is not known [31] and the discussion needs to be redirected accordingly.

In total, we carried out 16 interview sessions in an environment which we expected to be fruitful for MAR services, shopping centres. Shopping centres as a context embodies a diverse set of user needs and everyday tasks in which AR could be utilised: for example, acquiring additional product information, AR-supported navigation, mobile payment, advertisements and social interaction. In such contexts, people can have suitable needs as well as the required time and interest to use new technologies for either transient activities or long-term entertainment purposes. Furthermore, shopping centres can provide various services for different user groups and rich contexts in which to utilise user-created AR content. The sessions were arranged in two major shopping centres in Tampere, Finland, with eight sessions in each. Each venue has roughly 20−10020-100 specialised shops and services and is situated in an area with a high density of population.

To add further authenticity in the particular context, the participants were recruited to participate in different social settings: as single participants, in pairs and in groups. With

such diversity in the social setting, we aimed at covering both personal expectations related to single user interactions and group and collaboration perspectives on utilising MAR. This not only allowed us to facilitate rich discussion around a multifaceted theme but also to elicit requirements and expectations in different probable and authentic social settings. After all, shopping centres are places where various technologies are often used collaboratively (e.g. the use of mobile phones among friends) or, at the very least, in a socially vivid context.

2.1 Contextual interview session procedure

First, the participants filled in a simple questionnaire about demographic information (age, gender, education) and their general attitudes towards technology and sharing information (Table 1). Next, the participants were given a brief concept level introduction to the general idea of mobile augmented reality with a few illustrations of digital content augmented on top of a real-world view. This was necessary to ensure that the participants understood the central principles of AR and thus were able to envision its possibilities and anticipate the experiences with such novel technology. This allowed looking into the initial reactions and acceptance of a new technology. In addition, the participants were encouraged not to be stuck on the limitations of present-day technology. The topic was discussed as a comprehensive service, which would consist of device(s), application(s) and information content; that is, MAR was representing a “role prototype”.

Each session lasted 50-70 min, during which the participant(s) were interviewed and encouraged to invent ways of using MAR services in shopping centres, related to their personal needs and activities that they normally would do there. In practice, this meant touring around the shopping centre"thinking aloud" about their needs and ideas evoked at that moment. The participants were interviewed during

Table 1 Participants’ general attitudes towards the use of technology and information sharing

Scale: Likert 1-5 (1: strongly disagree, 5: strongly agree)
and after the touring about their underlying needs and expectations. The participants were asked concrete questions related to themes of needs for additional information in the current environment, creation and publicity of the AR information, interaction with the information, elements in the context that would affect how one would use such services, their feelings while envisioning the future use of MAR, perceived benefits and advantages and perceived weaknesses and risks of MAR services.

Because of the somewhat complex and subjective nature of an experience, participants’ expectations of what kind of experiences would be evoked with AR were approached indirectly, instead of explicitly inquiring about their abstract expectations of the probable emotional reactions in their future use of MAR. To add further authenticity, the participants were given a trivial physical object to interact with (a plastic phone-sized container) as a mock-up representation of an “omnipotent” futuristic MAR device (see Fig. 2). In each session, there were 2−32-3 researchers, of whom one was the facilitator of the session and the others documented the discussions during the sessions and observed the possible interactions with the mock-up. Furthermore, an audio recorder was used to ensure detailed recording of the discussion.

2.2 Participants

Nine interviews were held with individual participants, five in pairs and two group interviews with four and five participants. Interviews of all three types were held in both locations. Altogether, 28 participants took part in the 16 sessions, 12 of them male and 16 female. Their ages ranged from 20 to 42 , with a median of 24 . All of the participants were Finnish. In general, the level of education was high, over half of the participants either having a university degree or currently studying at a university. Most participants were familiar or very familiar with the environments in question, but there were also a few participants who had visited the location only a few times. Hence, we were able to address different kinds of needs and activities in terms of frequency in the specified context.

In the recruitment of the participants, they were required to be at least somewhat interested in new technical products and thus could be considered as early adopters of future MAR services. This was necessary in order to get participants who could quickly understand the main principles of MAR and would thus be able to express thoughts that are relevant, particularly in this technology domain. To judge from the background questionnaire, the participants generally had rather positive attitudes towards technology (see Table 1 for details). For example, most considered technology to be useful in their everyday life, but at the same time, they were not the first ones to acquire new

Fig. 2 Left contextual interview participant interacting with an imagined “omnipotent” MAR device. Right a participant envisioning the future use of MAR in front of a café

technology. Their general willingness to participate and contribute to social online services was on a level slightly below average, as was their willingness to share their location information. All in all, most of the participants can be considered to be potential early adopters of future MAR services-albeit not quite as “innovators” (the first ones to adopt new technology). Considering our qualitative research approach, the details about the participants’ backgrounds were not used as variables to carry out comparisons or further categorisation. The data were gathered merely to describe the potential users involved in the study.

2.3 Analysis

The qualitative data based on interviews and observations were analysed with a physical affinity diagram, which is a commonly used group analysis method for unstructured or semi-structured data from interview and observation studies [4]. First, on the basis of group analysis by four researchers, the transcribed recordings from the sessions were transformed into over one thousand affinity notes (i.e. individual items of analysis extracted from the interview transcriptions and observations). Most notes consisted of interpretations of participants’ expressions in the interview transcriptions (e.g. “the user wants to see a product tangibly before making a decision”), and some notes were direct quotes from the participants (e.g. “I would like to use the information more or less unnoticed”). Considering this and the fact that the quotes have been translated from Finnish, not all of the user quotes reported in the results section are verbatim. In addition to the research data, each note contained the session number to facilitate later traceability.

The interviews with the pairs and groups were each treated as one session: in the affinity notes, we did not differentiate between the members of the group or pair but considered such participants’ expressions as the general expressions of the group. This could be done because the groups and pairs usually agreed on the various matters, and we did not aim to quantify the results (e.g. counting the numbers of affinity notes under each category). In building
the affinity diagram (see Fig. 3 for illustration), we objectively and systematically identified common themes and meanings from the affinity notes and inductively built up a hierarchy of themes, that is, consolidation. The affinity diagram was built in a group that altogether involved 5 different researchers.

As the diagram included a huge range of themes (some of them irrelevant for the purposes of this study), we continued the analysis by focusing on specific topics around user experience. With the help of an affinity diagram walkthrough and a grounded theory approach, we systematically re-examined the affinity diagram, taking a perspective on the characteristics of expected user experience and user requirements affecting the overall user experience. Each new label was then added to the diagram next to the affinity note(s) that elicited it and, in the end, categorised. In the results section, these aspects and categories that were identified are described in Sects. 3.1 and 3.2. The data were not quantified on the basis of the categories as it was not considered favourable to quantify such unstructured data from studies with single user, pair and group interviews. Furthermore, the overall goal was to understand the various elements of user experience with MAR services in general-not to enquire into any statistical differences between the categories. Therefore, the

Fig. 3 Roughly one-third (6/18 sheets) of the affinity diagram (showing both yellow and pink topics and the white affinity notes under them)

results provide a broad insight into the extent of the matters that were discussed rather than revealing the quantities of answers or observations in each category.

3 Results

To start with, we briefly describe the most central use cases and types of proposed AR information content that the participants’ brought out as needs in the shopping centre context. Overall, a great range of information needs was elicited. However, because of the focus of this paper, we only provide a summary of them to introduce the results in Sects. 3.1 and 3.2, where these aspects are discussed from a viewpoint of user expectations and characteristics of expected MAR user experience.

Most participants emphasised the need for services that would help in pragmatic everyday activities. The value of an MAR service was mostly seen to be in providing useful additional information related to places, people, public transportation or basically any momentarily relevant issues nearby. Such information could be acquired either in situ, that is, nearby the physical object in question, or remotely in advance, before actually visiting the area.

One of the most discussed needs was acquiring additional information about products (For example, details about the product’s background and contents, such as where and how ecologically or ethically it had been produced and detailed ingredients of food products). Instructions for use and social information (e.g. which other people have bought or used the product) were often mentioned as well. Such information was considered useful in determining the value of an unfamiliar product and in comparing different product options. “It would be useful to get detailed data about home electronics with the device” (male, 27).

Another significant set of information needs related to the overall environment and receiving an overview of it: locating the most important services like ATMs, toilets, elevators and exits, customer service desks, and viewing shops’ opening hours and lengths of queues in various services. “It would be useful to know when exactly the shops are open; best if already in advance, in order to be able to plan your agenda” (female, 24). “It would be great to easily explore what to buy and what kind of discounts there are” (male, 32).

Navigation was another much discussed activity in which to utilise AR. The needs included both navigation inside the shopping centre (and it is nearby area) to specific shops or services and navigating to certain products inside the shops. “I would like to know to which direction you get from each exit” (female, 24). “It would be nice to see the exact place of a certain product in a supermarket to aid in searching. Often, even finding the right shelf would help”
(male, 27). All in all, the MAR services could both provide new relevant information in situ and facilitate the activities carried out in the physical world (e.g. wayfinding), hence making it possible for people to concentrate on the relevant and more cognitively demanding or pleasing tasks.

Related to ubiquitous computing in general, many participants expected the AR services to be proactive or context aware to some extent. The MAR services could suggest personally interesting products or activities nearby or remind the user of things to buy in situ, when he or she is nearby such products. “It would be nice if the device would inform the user when passing by a personally interesting product. Additionally, it would be great to know where to get student discounts” (group of 4 women, 24-25 years). Proactivity was also expected as automatically informing about critical information related to temporary changes in the environment, such as renovations, unexpected opening hours or other critical things the user should be aware of.

Furthermore, some participants brought up needs related to more pleasurable and stimulating aspects of life, for example, augmented art and self-expression. Such content was discussed from various viewpoints: for example, artistic layers on top of the real-world view, augmented stickers, repainting walls, virtual graffiti and decorations virtually added on the real environments. Additionally, social aspects like locating one’s friends and the AR service serving as a platform for social interaction were much discussed.

3.1 Characteristics of expected user experience

In the following, we describe the participants’ expectations of MAR services from an experiential point of view, that is, what kinds of experiences were expected to arise especially with such services. To roughly structure the findings, we report the experiences in three topics of product features, according to Hassenzahl [16]: (1) MAR Service Functionalities: expectations of functionalities and software features that the service would provide; (2) Information Content of MAR Services: expectations related to the augmented content created, accessed, shared and used in MAR services, as well as its characteristics, such as origin, accumulation, and perceived quality and relevance; (3) Interaction with and Presentation of MAR: expectations of the way of interacting with the augmented reality, that is, the ways of controlling the service and required mobile AR devices and other artefacts (input), and visualising the AR information and affordances (output).

3.1.1 Expected experiences related to MAR service functionalities

Mobile AR services were expected to make the tasks in shopping centres, trading and related activities feel more

efficient: reducing trivial tasks and waiting times, easing information acquisition, and saving effort or money, thus facilitating accomplishment. Especially in shops, additional information about product features and backgrounds was seen to make it easier to decide between several options, lower the need to remember trivial details, and reduce the number of unnecessary or unsuitable purchases. “The services would make shopping decisions quicker and, more importantly, easier” (female, 28). Additionally, way finding in urban environments was considered to become easier with such pervasive and multi-modal way of interaction. “[Getting AR navigation aids] would remove excess wandering in large shopping centres” (female, 24).

The service making product suggestions or proactively informing about discounts nearby were expected to further ease and enrich the shopping experience. “When arriving to the mall, I’d be most interested in the special discounts from different shops” (group of 3 male and 2 female, 24-26 years). A few participants saw a possibility in receiving timely information about the contents of various shops without actually visiting them.

Novel functionalities like location-specific reminders or 3D models of products and the environment would augment the user’s memory and capabilities, thus helping in trivial everyday tasks like remembering to purchase certain goods. Such features could increase the sense of empowerment, the user’s feeling of having new tools and resources that make possible novel activities and interactions. “I would want to virtually try out the product in its intended place of use before buying, for example virtually place the thing in my home” (male, 21). “Seeing through the walls would be great, so that I’d see an approaching bus upfront” (female and male, both 23).

As a counterforce for empowerment, however, a few participants pointed out that in carrying out target-oriented and pragmatic tasks, one’s efficiency might decline because of the too captivating way of interacting and attention focusing on irrelevant stimuli in the service. “This service could slow me down as I could be stuck in browsing products” (male, 27 and female, 26). Furthermore, extended use was worried to blur one’s conception of what is real and what is digitally augmented. Some participants shared their fear of becoming too dependent on the service and its automation, leading to reduced thinking in the long run. “I would not want to rely on the service in too many things, so that, e.g., my memory would weaken” (male 23).

Partly related to efficiency, increased awareness and knowledge about one’s surroundings was identified as another central expectation. As MAR provides a pervasive and holistic interface to physical places and objects, the embedded, latent, digital information in the environment becomes more perceivable and explicit, which shapes people’s perception of the surround world. “With the
service I could become aware of other people’s experiences of the shops” (female, 23). A lot of discussion was related to expanding one’s mind regarding various products, for example, food products and clothes. The product information would be about, for example, viewing what other product options there are, what kind of opinions people have of it, how it could be used or prepared, or what is the ecological and ethical background of the product. “The service could automatically compare prices between your favorite shops and tell where to get for example the cheapest cheerios” (male, 27). “We’d want the service to provide us with information that cannot yet be found on the packaging” (male, 23 and female, 22).

A few participants saw that MAR services could provide real-time notifications of relevant changes and happenings (e.g. altered opening hours, special offers, public events), thus boosting the momentary awareness of the current environment. “Information about busses being late and new arrival times would be a nice service” (male, 21 and female 20). In addition, social awareness came up in the interviews: for example, who of one’s friends or what kind of people in general are nearby, or viewing anonymous statistics of what kinds of people use a certain product might give the users a cue of how suitable the product might be for them. “Other users’ opinions are important in order to prevent marketing info to excessively affect product choice” (group of 2 female and 2 male, 24-25 years). Finally, a few participants contemplated whether with context-aware features and logging of the user’s behaviour it could be possible to review-and possibly change-one’s behaviour in the long run (e.g. what healthy or ecological products one consumes, how money and time are spent).

More hedonic and emotional elements like inspiration and surprise were also brought up with high expectations. “Such services would encourage me to try new things” (female and male, both 23). The high expectations mainly based on participants’ assumptions of the service being aware of the user’s context and earlier behaviour and having proactive features. “Getting suggestions from the service would be nice after setting some basic preferences, for example, when looking for a present for someone” (female, 28). A few participants mentioned that information about nearby public events and happenings within a few hours would be very intriguing. “In some situations, a ‘what could I do today?’ -button would be nice to get ideas” (female, 26 and male, 27).

Location-aware AR services were expected to proactively offer positive surprises by providing the users with information that is relevant or interesting in the current situation and location. For example, this came up as ideas for getting suggestions to try out products that are favoured by other people with similar interests or values. "The device could notify by voice or vibration near a pharmacy if you

lack some product you need: ‘remember to buy it now!’" (female, 23). Such proactive and surprising features would support the joy of finding as well as curiosity towards new services, products or other aspects even in familiar environments.

Furthermore, experiences of surprise and inspiration could also originate from the information content and social characteristics of the services; other people create and share their own content and share opinions, thus expanding the amount of information and options related to a certain target. “I would like to get surprised, that’s why especially the information created by people I don’t know would be meaningful” (male, 23 and female, 22). In addition, the versatile and social nature of MAR could increase appropriation of the services in various ways (i.e. adapting and repurposing the technology for new ways or purposes of use [44]). For example, several participants shared an interest of creating personalised augmented views of the shopping centre, e.g., by making the environment more cheerful with decorations.

When inquired, many were interested in an idea of having “hidden” AR content that could provide pleasurable surprises, for example, with special offers in the shopping centre area that could be accessed only with the help of MAR. One participant (male, 27) shared his idea of people being able to act as AR advertisements by highlighting the brands of clothes or other garments when they appear in other users’ augmented views. Another one (female, 26) envisioned that the surroundings could also be used as visual platforms for AR games, art or other entertainment.

As a downside, some participants mentioned that inspiring features and giving suggestions could also lead to impulse purchases or other dubious behaviour, because of people focusing on the pleasures of the moment and acting based on ad hoc needs and AR stimuli from the environment. “I could try something totally new based on the information provided by the service” (male, 27).

Overall, MAR services were expected to inspire, surprise and encourage people, for example, by letting their creative characteristics come out or by offering something extraordinary and stimulating. “On a sad day the service could cheer me up with an indulgence mode” (male 27 and female 26). The pervasive and immersive nature of AR visualisations can be seen to further catalyse such positive experiences. The importance of elements of surprise and inspiration are emphasised in their ability to maintain users’ interest in the service over time.

3.1.2 Expected experiences related to MAR information content

With regard to MAR content, the participants expected that experiences like collectivity and connectedness would be
evoked. For most, it was self-evident that the information content would be socially constructed and enriched, largely based on users’ self-created content. “The information shared by users would create a sense of community” (male, 26). AR was expected to serve as a fruitful tool for content sharing, self-presentation and digital community building. Therefore, various social experiences would emerge in collective and collaborative use of the service content and creating its content, in utilising the socially aggregated AR information in personal use, as well as in using AR in facilitating social interaction.

For example, several participants mentioned that they would be interested knowing when familiar people appear nearby, for example, by highlighting familiar people in the augmented view. “When arriving to the area the service could tell me if there are friends there” (male, 32). Furthermore, one visionary participant saw that MAR services could provide totally new ways of social interaction and a new tool for people to communicate their identity. “With the service I could organise or view location-based gatherings of various groups that I’m interested in” (male, 32). “The MAR application would not only be a device but a digital version of the user-an extension of one’s current reality” (male, 32).

The most commonly mentioned type of social information was other people’s recommendations about different products and services. “I’d be very interested in other users’ comments, especially the local people’s comments while abroad” (female, 28). It would also be useful to know whether anyone familiar has bought a certain product one is interested in. Statistics about how many and what kinds of people have bought the product in general were seen useful as well. The reviews and information shared by other users, especially by familiar people, were often considered more relevant than those created by service providers. “I regard friends’ ratings as more important and significant than those by ‘just anyone’” (female, 24). Reciprocally, users sharing their opinions or further commenting existing ratings would facilitate the sense of participation. “I would give product feedback in any case, bad or good” (female, 42). “A great advantage in AR services is that they could increase the power of consumers” (male, 26).

Closely related to the above mentioned is the experience of liveliness derived from the expectation of the content being constantly updated and changed by the service providers and other users, as well as aggregated from other services and domains, such as the Internet. Content was expected to have a certain novelty value, that is, not only being updated explicitly by user- and community-created content but also implicitly by the users’ actions in the service and automatic logs and traces from them. “I would leave tags and comments with the service” (male, 32).

Consequently, the services and environment might seem positively dynamic, vivid and inspiring but on the other hand hectic, confusing and unstable. Such liveliness was also expected to facilitate sharing further content and ITmediated social behaviour, such as collectively created MAR art pieces or finding like-minded people (e.g. those who visit same places).

In addition to providing practically and socially beneficial services, mobile AR was expected to offer playful and entertaining momentary experiences. Gaming-like and playful services, or aspects in them, can further catalyse hedonic experiences, such as self-expression, enjoyment and stimulation. In general, artistic and amusing content was mostly considered unnecessary or trifling. Several participants, however, emphasised that the service should also be playful, fun and exciting, and catalyse creativityat least when one has extra time to spare. They brought up ideas of creating personalised, humoristic and artistic views of otherwise dull environments, thus gaining a way to locally express oneself. “Sharing and creating art and other self expression would be interesting especially in foreign surroundings” (female, 23).

Such playful content was discussed with various perspectives: artistic layers on top of the real-world view, augmented stickers, repainting walls, virtual graffiti and decorations virtually added on the real environments. “I would love to draw and paint extra content in shopping centres and share it with others” (female, 26). “In creating content myself, I could try various things and perform some foolery. It could also be easier to create art as augmentations than by other means” (male, 23 and female, 22). Creating also social implications, funny digital content left in the environment was expected to evoke social play, such as treasure hunt or creating location-based secrets within groups of people. “[MAR] gaming applications would be nice! Such would work well in my group of friends” (female, 24). Overall, including playful aspects is a fruitful way of introducing the user to new technologies, types of interaction or services, thus facilitating successful adoption of them. In MAR, mixing digital content with real-world objects offers an encouraging tool for enhancing customer commitment both by service providers and users themselves.

3.1.3 Expected experiences related to MAR interaction and presentation

Overall, interaction with the required devices was rather little discussed-most probably because the study set-up did not allow concrete human-technology interaction. Nevertheless, most participants expected that most of the augmentations would be in a visual form: icons, simple symbols text, or layers of colours, etc. “The augmented
information in such a busy environment would mostly need to be visual” (female, 24). In addition to this, other interaction modalities than audio were not brought up. Furthermore, except for a few mentions of data glasses and visual projection, most interaction was envisioned to be based on the mobile device screen. “The AR information should be able to be viewed also from a bigger screen than the mobile device, for example when navigating”(male, 23 and female, 20). “It would better to project the map on the ground” (female, 20 and male, 23). Data glasses were considered most suitable in intensive and long-lasting interaction where comprehensiveness of interaction with the AR information and elements of efficiency and increased awareness are central. On the other hand, glasses were seen be useful with automatic service-initiated augmentations where the elements of surprise and inspiration are important. “With data glasses I could see the entire environment, and hands would be free for other things” (male, 27).

AR being rather novel as a concept, interacting with the augmented reality was nevertheless expected to be intuitive and easy to learn. Most participants identified well with ideas of pointing and shooting (magic lens metaphor [42]) or touching objects in the real world. They considered such interaction as a very natural way to interact with location-specific digital information, especially in ad hoc and short interaction situations (e.g. taking a snapshot of the augmented view, and interacting with the still image). “Pointing the products [to gain additional information] would be very convenient, if it really worked” (female, 28).

In addition, aligning digital information directly on top of the real-world counterparts was considered most natural as an interface. The ontology and hierarchy of information, objects and metadata in AR are based on real-world objects and places, which lowers the ambiguity and helps understanding to what target MAR information is related to. It seemed natural for participants to think that information is related to a certain real location or object, and thus the related information could also be accessed by interacting with it. “Getting additional information should be based on touching, e.g., products with the device” (female, 26 and male, 27). Furthermore, because of the extent of AR content related to various aspects and the novel way of interacting with it, the MAR services were expected to be captivating. “The application would probably slow people down as they stopped to wonder the new information” (male, 27 and female, 26). Augmented reality could make the digital content feel more immersive and tangible when compared to conventional mobile and Internet services. “I could project the AR view to a wall and by doing so get more information; it would be kind of tangible” (female, 42).

3.2 Design requirements for pleasant MAR user experiences

This section further describes expectations that relate to the service features and contextual issues-instead of directly representing certain experiences. In other words, these can be determined as design requirements for mobile augmented reality services. The realisation of these in a MAR service would affect the quality and valence of the user experience in general and affect whether the above-mentioned experience categories can come true; that is, when realised appropriately, these allow the positive and delightful experiences to arise, but when realised badly, these might prevent a specific experience from arising.

3.2.1 Requirements for MAR service functionalities: privacy and control, and reactivity

The needs for privacy management, explicit user control and feeling of having information secured were emphasised in most interviews. The participants were concerned about what information about their activity will be saved and where, how public is the interaction with the service, and who can eventually access the content they have shared themselves. “The information about user that has been saved on the device must not spread anywhere” (female, 23). All of these are common challenges in design of almost any information systems. However, MAR and context awareness as new instruments and platforms for creating and sharing information and interacting with it in various mobile contexts add further aspects to these topics.

The levels of sharing self-created content consolidated earlier findings: “There should be different options in sharing information, some of it shared publicly and some only with friends” (group of 3 males and 2 females, 24-26 years). Regarding user control, a 26-year-old male mentioned “I’d want the device to be totally under my control. The device automatically learning about me would be too disturbing”. Additionally, the device might “know” too much about the user: the more the service logs the user’s activities, the more critical the risk in leaking information is. “After a while, the device would know very much about me, which kind of disturbs me” (female, 24).

Despite the general positive acceptance and its stimulating nature, the interaction with AR received doubts related to its social acceptance in such contexts with great number of people. A few participants regarded using the MAR services too conspicuous in public places as well as risky with regard to the privacy of what information one is accessing. “I would definitely like to interact with such a device in an unnoticeable way” (female, 24).

The fundamental human need for physical safety was manifested as risk of being too greatly immersed in the

AR service. In busy environments, one might, for example, not notice and, therefore, collide with, moving physical objects in the surroundings. “One’s focus being drawn from the real world is definitely a risk” (group of 3 males and 2 females, 24-26 years). In other words, the service has to be aware of, and reactive to, the physical risks in the current environment and the user’s possibilities to react to the service’s output. “There should not be too much information on the device screen so that the user’s attention can be focused on the real world as well” (female, 24).

3.2.2 Requirements for MAR content: relevance and reliability

The most discussed user requirements were the demands for relevance and personalisation. The provided information content was expected to be useful, emotionally important or otherwise personally meaningful in the current situation. “The level of detail that you need depends totally on the situation and my current needs” (female, 28). As a result of the assumedly immersive and pervasive AR interface, many were afraid of gaining too excessive amount of information and thus not being able to concentrate on interacting with the real world, nor the augmented information. “The biggest risk could be flooding of information, which could lead to missing something important and growing numb with the ARA R content” (male, 21 and female, 20).

Relevance was considered salient especially when the information is automatically “pushed” to the user (e.g. advertising or notifications) and the amount of information is high. “I’m ok with ads if they are personalised to correspond to my needs” (female, 23). Despite the high expectations of the service’s awareness of the user and context, its effectiveness was also questioned. “I wonder how the service could know what I like each day” (female, 24).

The participants brought out various ideas to filter unnecessary content and ensure its relevance. For example, they mentioned content classifications, setting user profiles, modes and utilising information filters during the use of MAR service, as well as teaching the service by giving feedback during the use. “The personalisation could work like this: first showing the kind of info I normally look for and the other kinds only after that” (female, 24). User profiles could, for example, determine how extensively and what type of information is filtered by default. “I’d gladly create a profile of myself, based on which the device would then filter the content” (male, 27 and female, 26). A possibility to instantly show either appreciation or disinterest to the provided content (teaching the service one’s preferences) and being able to add bookmarks and ratings to browsed content were mentioned as essential features. "Delivering advertisements with the MAR app should work

so that when the I remove a certain ad few times without reading it, the ad stops appearing to me" (male, 32).

Another critical expectation was reliability of the content. It refers to how tenable, up-to-date, and valid the offered information is perceived, thus implying how much it can be trusted. Validity of the information was questioned especially in such everyday situations where information needs relate to things that tend to change often (e.g. daily lunch menus, temporary offers, exceptions and changes in environment). “The information should be up-to-date with accuracy of at least one day” (male, 27).

Most participants would trust the validity of content provided by public and official institutions and well-known vendors more than that created by other users. User-created content was not seen very reliable and trustworthy as it would most probably be based on certain active users’ viewpoints and, for example, comments of exceptional experiences with products or services. “I would give feedback only if I was either very satisfied or very unsatisfied with the product” (female, 23). Therefore, several participants emphasised the need for knowing who or what organisation has created and shared the content (e.g. number of people who have used or contributed to creating it, and what kind of users they are in contrast to oneself). “The more there are comments, the more reliable is it” (male, 27 and female, 26).

3.2.3 Requirements for interaction: easy and flexible access, distinct affordances

Due to the conceptual level approach in our research, concrete user interface level aspects were discussed rather little. Nevertheless, the participants highlighted various usability-related issues, such as straightforward access to AR information: it was expected to be accessible instantly when viewing the service. In addition, flexibility and control in interaction were much emphasised. “The automatic content should be easily removable, e.g., with a simple sweep on the screen” (male, 23 and female, 22). “I would not want to hold the device in my hand all the time while shopping” (male and female, both 23). In design of MAR interfaces, the balance between allowing immediate browsing of information related to the real world and avoiding information overflow at the same time can become a challenge as the contexts of use and amount of available content probably vary radically.

Delivering the affordances (i.e. the action possibilities latent in the environment [14]) was identified as a challenge when the options and digital information related to the physical environment are abundant. “Information should first appear as some kind of a visual hint over the viewed object, through which the user could then view more detailed information about it” (male, 32). As people
move about in the physical environment, the services cannot rely on having the user’s undivided attention. “[In middle of all the other content] the navigation guides should be very clear and noticeable” (female, 22 and male, 23). Cues about what information is available should be subtle and sensitive to the context (e.g. haptic or auditory cues for the most urgent information and visual for less relevant information). Additionally, the physical counterpart on top of which something is augmented should be visible from the user’s location. “It would be best to get info only about those things that I can see, and the augmentation should be pretty precise” (male, 27).

4 Discussion

In this section, we first summarise the central user expectations, which are followed by an analysis of the most central technical enablers that seem to matter in the expected UX of MAR services. Finally, we discuss the validity of the research approach and methodology.

4.1 Expected UX in mobile augmented reality

Overall, the expectations related to various elements of the service, such as functionalities and software features, information content, interaction and elements of the context of use. Hence, the user experience of future MAR services can become very multifaceted and affected by an extensive set of design aspects and technologies, as well as the user’s internal expectations, concerns, attitudes and moods [6, 18]. This study addressed this research challenge by exploring the user experience prior to actual use by focusing on potential users’ anticipated experience of the envisioned MAR services. The categories regarding both the expected characteristics of experience and design requirements (experience enablers) are summarised in Table 2.

The diversity of the expectations that were identified is noteworthy when reflecting the expected experiences to earlier UX frameworks and theories that discuss various types of human experience created in interaction with interactive products. Considering the framework by Hassenzahl et al. [19], we can conclude that the characteristics of user experience and user requirements that were identified were mostly related to pragmatic and utility-oriented aspects. For example, the participants expected MAR services to catalyse experiences of efficiency and empowerment, increased awareness and knowledge, intuitiveness, and required them to be usable and offer relevant, personalised and reliable content in a privacy-sensitive and safe way. Nevertheless, emotional and hedonic characteristics of experience were also anticipated. For example, inspiration, liveliness and surprise are related to both

Table 2 Summary of the user expectations of MAR: characteristics of expected user experience and requirements that affect and facilitate the experiences

stimulation and evocation, whereas captivation and playfulness relate mostly to stimulation, and awareness, collectivity, connectedness and personalisation mostly to identification. Such emotional expectations are important to consider in order to exceed the user’s pragmatic expectations and provide additional emotional value and a pleasurable experience [6,19][6,19].

With regard to the components of product experience of Desmet and Hekkert [5], the characteristics of experience identified in this study also cover their categories rather well. For example, aesthetic experiences are represented in our categorisation by captivation, inspiration and, to some extent, by playfulness and liveliness as well. Experiences of meaning or symbolic significance can be identified in connectedness, empowerment and the requirement for relevance. The emotional experiences are manifested here, for example, as surprise, inspiration, playfulness and liveliness. This wide theoretical coverage is by no means to
claim that the diversity of UX in MAR is now fully understood and no other characteristics of user experience would be relevant in MAR. On the contrary, understanding how these categories are realised in actual use and what the actual relative importance of each category in different contexts is requires further research.

4.2 Building blocks of UX in mobile augmented reality

The comparison of our findings with the existing UX frameworks indicates that the UX is particularly complex when discussing futuristic and widely applicable technologies like MAR. The users had high expectations of MAR and how it could help them in various activities, make possible novel interactions and provide relevant information in situ.

Considering the specificity of the results in this technology domain, we need to remark that some of the

expectations also relate to technology domains other than purely mobile AR. We could question how well we were able to identify the most MAR-specific expectations, and how novel they are as user research findings. For example, expectations of personalisation, reliability of content, intuitiveness or empowerment can be seen as somewhat generic aspects that are often expected or required from new technology, especially with regard to mobile and ubiquitous computing.

However, we must bear in mind that the aspects highlighted in the results represent the potential users’ main expectations and concerns of AR services, that is, holistic entities that also include other features than the AR visualisation per se. The participants regarded MAR services as comprehensive, comprising various functionalities, information content and ways of interacting with the content. Some expectations were also attributed to the activities and contexts that such services could support and the participants easily conceptualised that MAR would inherently encompass aspects such as context awareness, proactive features and personalisation of content. Thus, MAR services are also burdened with expectations that are not necessarily specific to AR as such but relate to interactive technology or ubiquitous computing in general. After all, for an end user, it is very challenging to assess the additional value of a particular technology alone, and therefore, it is difficult to elicit the most AR-specific expectations in the first place, as well as to distinguish them in the user data.

With further analysis and reasoning, we can nevertheless distinguish four different components (i.e. “building blocks” or “layers”) of enabling technology that mattered to the participants and to which the expectations were mostly directed to in such pre-use research setting: (1) AR as output technology, (2) embedded information, (3) con-text-sensitivity and (4) mobility. Considering design, it is important to understand how the expected experiences could be formed in interaction with the service and what aspects in the service contribute to them. These different components of the service can enable or disable an experience (make it possible or impossible to happen), evoke or initiate it, and catalyse or stimulate it or its valence and strength. The overall MAR user experience can be seen as a combination that is based on these different cornerstones of the technological fabric of MAR.

First, some of the expectations can be considered to relate mostly to the ARA R technology, that is, the characteristics of augmented reality as system output and the way of interacting with the mixed reality information. The additional value from augmented reality seems to relate mostly to experiences of captivation and intuitiveness (i.e. those related to interaction with AR) but to playfulness, inspiration and creativity, too. AR can effectively evoke such experiences because of its visual and immersive way of
visualising information and by providing the user with a very immediate and intuitive way to interact the augmented information. AR can holistically change the way of viewing the surrounding environment and interacting with related information. Aligning the layers of digital and real information on top of each other creates an intuitive and immersive interface to one’s digital surroundings.

Additionally, the general novelty value that AR currently has can be considered to explain some of the expectations towards having a captivating and inspiring user experience. Furthermore, AR can be seen to effectively catalyse, boost or nourish other experiences; such that perhaps result from other components of technology or service features, but AR catalyses the experience or increases (or decreases) its strength. For example, increased awareness, efficiency and liveliness would gain from AR. It provides countless possibilities to take advantage of the new digital layer of information in various everyday purposes. Considering the design requirements (Sect. 3.2), the expectations with regard to interacting with the AR content were high, which creates challenges for accuracy, timeliness and robustness of the AR information visualisation.

Second, the information content discussed in the interviews (e.g. about products and services in urban environments) largely relates to general ubiquitous computing technologies and concepts like ambient intelligence [57] and the internet of things [8]. That is, real-world objects would be embedded with rich information created both by service providers and end users or aggregated from other services. To support users in their everyday tasks, such “intelligent” objects would be networked together, make themselves recognisable and be responsive to the presence of people. Mobile AR can be seen as a potential interface to browsing such information related to real-world targets, which was evident in the interviews as well. In that sense, this component of enabling technology can be seen as a necessary enabler for experiences like efficiency, empowerment, increased awareness, liveliness, captivation, surprise and collectivity. Without technologies to integrate the digital content to physical objects, the usefulness of AR would be very limited, especially in the contexts of use discussed in this study. The expectations related to content (reliability and relevance) are naturally such that are mostly directed towards this component.

Third, some experiences could originate mostly from utilisation of context and location sensitivity: the service’s functionalities and content being determined by and adaptive to the current context (e.g. location, social surroundings). Proactivity, which was often expected from MAR services, can also be said to relate to this component. Such adaptivity can be said to contribute especially to experiences of increased awareness and surprise, but also to efficiency and empowerment.

Location awareness brings additional value to the AR content by enabling more contextual and location-based relevance. Naturally, relevance is a central requirement also in other domains, such as in Internet. However, this expectation was emphasised because the AR content was expected to be related especially to a certain location, moment or object in the user’s surroundings. Additionally, as relevance of information might quickly expire due to changes in physical context (e.g. user movement, change in social context), MAR services were expected to be sensitive even to minor changes. Surprise can be a common element in the use of various technologies, but proactivity and context-sensitivity can be seen to add further aspects to it. Experiences of surprise were often mentioned to be related to the location where the information is received, that is, how relevant and temporally interesting a proactively prompted content is in the situation. On the other hand, AR can be seen as a powerful interface in utilising such proactive features and bringing the content instantly at the user’s disposal.

Finally, an experience can be a result of mobility. That is, the experience is common and possible in most mobile services or even in interactive technology in general. Such experiences can be based, for example, on the characteristics of mobile technologies (allowing authoring of content and interaction with it anywhere, anytime) or new types of interaction used in mobile devices and mobile contexts in general (e.g. speech and gestures). For example, social experiences like connectedness and collectivity, and liveliness can be seen generic to almost any interactive technologies or Internet services where users create part of the service content.

Social features and user-generated content being wide spread in modern Internet services, it was hardly surprising to identify similar expectations related to social characteristics in MAR services too. In addition, privacy and reliability of content have often been found as critical elements in mobile and ubicomp studies and thus have been extensively discussed in related literature (e.g. [28]). As with any technology, the privacy burden that the user takes must be balanced with the benefit gained. The amount, exactness and sensitivity of information the user is willing to share with the service depends on what is the value gained from the feature that requires the user’s information. Nevertheless, location awareness and AR as system output were also expected to cause new privacy risks to consider. For example, as the participants were concerned about revealing their location information, it should be enough to reveal only the approximate location instead of precise GPS coordinates. Most AR information was expected to be accessible without logging in anywhere, sharing any personal information or letting the service log one’s activities. In addition, privacy is important in
creating a private AR interface instead of, for example, an interface where AR information is projected on physical surfaces and thus accessible by surrounding people.

4.3 Re-examining the approach

In retrospect, the research approach naturally had some drawbacks and left some aspects of user experience disregarded. Studying expectations of technology and services that do not yet exist is unquestionably a challenging task.

First, although the descriptions of the idea behind MAR were intended to provide a general understanding of MAR as a concept and thus facilitate discussion, it is likely that they had a “priming effect” [51]. For example, the participants focused mostly on visual AR content, and, for example, audio AR was not much discussed.

Second, the participants were mostly representing the Finnish culture, which might have affected what kind of experiences people expect of technology, and what aspects are considered to affect the overall user experience. The participants were mostly technically oriented, which became evident with the generally positive attitudes towards new technology and services and thus affects the generalisability and transferability of the results to other populations. Nevertheless, such sampling-related limitations are often inevitable in qualitative research, and therefore, conclusions should be made only in the light of the limitations.

Third, with such a pre-experience research approach to metaphorical or anticipated experiences, we could not deal with the actual experiences created in interaction with real, fully functional MAR services. The expectations reported here can have been influenced by the participants’ earlier experiences of interacting with the real world [47] and existing information systems, such as web sites, virtual worlds or other mobile and location-based services. When further reflecting to prior UX literature, we see that, for example aesthetics, ergonomics and other specific UI/interaction level aspects were not much brought out in the discussions (for example, demands for good readability of the AR content, proper feedback from the system, accuracy and timeliness in aligning the augmentations, or audio-visual authenticity of the AR content). The research approach was too intangible and hypothetical for participants to consider such issues. However, after having identified such methodological limitations in advance and compared them to the benefits, the realised study set-up was a deliberate choice.

Finally, people are often incapable of considering the long-term implications of using a technology. The experiences discussed in this paper can be expected to be the initial “first-hand” experiences resulted from interacting with MAR services, thus being positively affected by the novelty value (e.g. temporary “WOW”-effects). Therefore,

we could not look into MAR-specific user experience aspects related to, for example, trust, product attachment or other long-term elements of UX. Studying such aspects would require a longitudinal research approach with a functional service. Additionally, it is too early to predict whether the reported experience categories are significant over longer periods of interaction (e.g. over the product use lifetime), and how the user experience changes as people appropriate the services to their own contexts and purposes (e.g. self-expression).

5 Conclusions and future work

The contribution of this paper is the new understanding of potential end users’ expectations of the user experience with MAR services. Understanding expectations is crucial as they reflect the anticipated behaviour, influence the user’s perceptions of the product’s capabilities and quality and most importantly, set a baseline against which the overall goodness of the service becomes assessed when actually using it. Being based on expectations, the experience categories can serve as targets to aim at in the design of MAR services. Additionally, from the participants’ expectations, we have identified the central design requirements (requirements for functionality, interaction and content) and technological components that seem to play a role in creating such experiences. Although with this study, we can provide only tentative design considerations, identifying these aspects serves as a basis for further research and exploration of how the UX builds up in the technology domain of MAR. For example, the details in the descriptions of expected experiences can be used as a basis for creating metrics for measuring the actual user experience as felt.

Based on the participants’ expectations, we can conclude that user experience of mobile augmented reality services is expected to be very multifaceted and affected by various components of underlying technology. The participants directed a rich set of expectations towards such services, ranging from proactivity, relevance and context-sensitivity to social, surprising, immersive and inspiring elements. Despite the inevitable shortcomings in the methodology, we regard that the results provide rich insights into the expectations of potential end users of MAR.

Because of the pervasive and engaging nature, and thus having certain novelty value, MAR possesses a great potential of facilitating positive user experience and emotions. By meshing places and times with digital content, mobile AR services could provide the user with information that is contextually very relevant, captivating or entertaining in the exact moment and location. AR content was expected to offer rich, lively and context-related information for ad hoc needs in mobile contexts, thus providing the
user with delight and surprising experiences. Personal relevance and personalisation of the service are important because AR embodies very immersive and personal aspects: it changes the very way of how one perceives the world and defines what information the user is able to receive about the environment (i.e. augmenting the human sensory system) and what meanings to attach to it. As a consequence, MAR has the potential of empowering people with new capabilities and facilitating people’s understanding of the surrounding world and the information related to it.

With regard to future research, we aim at systematically conducting further research on expectations in various application areas in order to establish a more comprehensive framework of the expectations of MAR services. After the most crucial expectations for user experience have been identified, it is equally important to explore how to reach them by design solutions. The role of expectations needs to be assessed against actual experiences with functional MAR demonstrators, for example, with longitudinal study set-ups and in authentic contexts. It is still to be seen how the longterm UX with MAR evolves, and what kind of personal and societal level implications AR technology might have on users’ perceptions of one’s surroundings, ways of running day-to-day errands and behavioural tendencies.

In addition, the question of what are the most AR-specific experiences (i.e. affected or evoked because of the AR technology instead of the other technological layers) remains to be further researched. As UX research in general is more or less in its infancy, a generally interesting question for the UX community is to what extent is it possible to distinguish domain-specific UX characteristics in the first place? Recently, we have been continuing the research on users’ expectations with other methods, for example, online surveys (e.g. [37]) and long-term user diary studies, as well as by leaning on different theories and frameworks around the concept of UX. Our next steps include operationalising the expectations into more specific metrics to be used in the evaluation of the MAR demonstrators developed in our research program.

Acknowledgments The research was supported by the Finnish Funding Agency for Technology and Innovation (TEKES) under the project DIEM (grant 40223/07). We highly appreciate the contribution of all the participants. For insightful comments on early paper drafts and fruitful collaboration, we thank Pirita Ihamäki, Tomi Haustola, Minna Kynsilehto, Timo Partala, Leena Ventä-Olkkonen and Tuomas Vaittinen.

References

1. Arhippainen L (2009) Studying user experience: issues and problems of mobile service-Case Adamos: user experience (im)possible to catch? Dissertation, University of Oulu

2. Azuma RT (1993) Tracking requirements for augmented reality. Comm ACM 36(7):50-51

3. Barba E, MacIntyre B, Rouse R, Bolter J (2010) Thinking inside the box: making meaning in a handheld AR experience. In: Proceedings of ISMAR 2010, Seoul, Korea, IEEE, pp 19-26

4. Beyer H, Holtzblatt K (1998) Contextual design. Defining customer systems. Morgan Kaufmann, California

5. Cheok AD, Yang X, Ying ZZ, Billinghurst M, Kato H (2002) Touch-space: mixed reality game space based on ubiquitous, tangible, and social computing. Pers Ubiquit Comput 6(5-6):430-442

6. Desmet P, Hekkert P (2007) Framework of product experience. Int J Des 1(1):57-66

7. Desmet P, Porcelijn R, van Dijk M (2007) Emotional design; application of a research-based design approach. Know Technol Pol 20:141-155

8. Dodson S (2003) The internet of things. The Guardian, Oct. 2003, http://www.guardian.co.uk/technology/2003/oct/09/shopping.new media

9. Drascic D, Milgram P (1996) Perceptual Issues in Augmented Reality. In: Proceedings of the stereoscopic displays and virtual reality systems III, SPIE, pp 123-134

10. Dünser A, Grasset R, Billinghurst M (2008) A survey of evaluation techniques used in augmented reality studies. In: Proceedings of ACM SIGGRAPH ASIA 2008 courses, Singapore, 2008, Article No.: 5

11. Dünser A, Grasset R, Seichter H, Billinghurst M (2007) Applying HCI principles to AR systems design. In: Proceedings of MRUI’07 workshop at the IEEE virtual reality 2007, Charlotte, NC, USA, pp 37-42

12. Feiner S (1999) The importance of being mobile: some social consequences of wearable augmented reality systems. In: Proceedings of IWAR '99 (International Workshop on Augmented Reality), San Francisco, CA, USA, pp 145-148

13. Gabbard J, Swan E (2008) Usability engineering for augmented reality: employing user-based studies to inform design. IEEE Trans Vis and Comput Graph 14(3):513-524

14. Gibson J (1977) The theory of affordances. In: Shaw RE, Bransford J (eds) Perceiving, acting, and knowing. Lawrence Erlbaum Associates, New Jersey

15. Gleue T, Dähne P (2001) Design and implementation of a mobile device for outdoor augmented reality in the ARCHEOGUIDE project. In: Proceedings of the 2001 conference on virtual reality, archeology, and cultural heritage, Glyfada, Greece. ACM Press, New York, pp 161-168

16. Hassenzahl M (2004) The interplay of beauty, goodness, and usability in interactive products. Hum-Comput Interact 19(4): 319−349319-349

17. Hassenzahl M, Diefenbach S, Göritz A (2010) Needs, affect, and interactive product-facets of user experience. Interact Comput 22(5):353-362

18. Hassenzahl M, Ullrich D (2007) To do or not to do: differences in user experience and retrospective judgments depending on the presence or absence of instrumental goals. Interact Comput 19(4):429-437

19. Hassenzahl M (2003) The thing and I: understanding the relationship between user and product. In: Blythe M, Monk AF, Overbeeke K, Wright P (eds) Funology: from usability to enjoyment. Kluwer, Dordrecht, pp 31-42

20. Herbst I, Braun AK, McCall R, Broll W (2008) TimeWarp: interactive time travel with a mobile mixed reality game. In: Proceedings of MobileHCI 2008. ACM Press, Amsterdam, pp 235-244

21. Häkkilä J (2006) Usability with context-aware mobile applications. Dissertation, University of Oulu

22. Höllerer T, Feiner S (2004) Mobile augmented reality. In: Karimi H, Hammad A (eds) Telegeoinformatics: location-based computing and services. Taylor & Francis Books, London

23. ISO, International Organisation for Standardisation (1998) ISO 9241-11: Ergonomic requirements for office work with visual display terminals (VDTs)—Part 11: Guidance on Usability

24. ISO, International Organisation for Standardisation (2009) ISO FDIS 9241-210:2009: Ergonomics of human system interac-tion-Part 210: Human-Centred Design for Interactive Systems

25. Kaasinen E (2003) User needs for location-aware mobile services. Pers Ubiquit Comput 7(1):70-79

26. Kuikkaniemi K, Turpeinen M, Salovaara A, Saari T, Vuorenmaa J (2006) Toolkit for user-created augmented reality games. In: Proceedings of MUM 2006, Stanford, California, Article No.: 6

27. Law E, Roto V, Hassenzahl M, Vermeeren A, Kort J (2009) Understanding, scoping and defining user experience: a survey approach. In: Proceedings of the CHI 2009. ACM Press, New York, pp 719-728

28. Lehikoinen JT, Lehikoinen J, Huuskonen P (2008) Understanding privacy regulation in Ubicomp interactions. Pers Ubiquit Comput 12(8):543-553

29. Lindgaard G, Dudek K (2003) What is this evasive beast we call user satisfaction? Interact Comput 15(3):429-452

30. Mackay WE (1996) Augmenting reality: a new paradigm for interacting with computers, La Recherche, Mar. 1996

31. Maguire M (2001) Methods to support human-centred design. Int J Hum-Comput Stud 55:587-634

32. Milgram P, Kishino F (1994) A taxonomy of mixed reality visual displays. IEICE Trans Inf Syst E77-D(12):1321-1329

33. Morrison A, Oulasvirta A, Peltonen P, Lemmelä S, Jacucci G, Reitmayr G, Näsänen J, Juustila A (2009) Like bees around the hive: a comparative study of a mobile augmented reality map. In: Proceedings of CHI 2009, Boston, MA, USA. ACM Press, New York, pp 1889-1898

34. Nilsson S (2010) Augmentation in the wild: user centered development and evaluation of augmented reality applications. Dissertation, Linköping University

35. Oliver RL (1977) Effect of expectation and disconfirmation on postexposure product evaluations-an alternative interpretation. J Appl Psychol 62(4):480

36. Olsson T, Ihamäki P, Lagerstam E, Ventä-Olkkonen L, Väänä-nen-Vainio-Mattila K (2009) User expectations for mobile mixed reality services. In: Proceedings of ECCE’09, Espoo, Finland. ACM Press, New York, pp 177-184

37. Olsson T, Salo M (2011) Online user survey on current mobile augmented reality applications. In: Proceedings of ISMAR 2011, Basel, IEEE

38. Oulasvirta A (2004) Finding meaningful uses for context-aware technologies: the humanistic research strategy. In: Proceedings of CHI 2004, Vienna, Austria. ACM Press, New York, pp 247-254

39. Peng ZR, Tsou MH (2003) Internet GIS, distributed geographic information services for the internet and wireless networks. Wiley, New Jersey

40. Remagnino P, Foresti GL (2005) Ambient intelligence: a new multidisciplinary paradigm. IEEE Trans Syst Man Cybern A Syst Hum 35(1):1-6

41. Rohs M (2007) Marker-based embodied interaction for handheld augmented reality games. J Virtual Real Broadcast 4(5)

42. Rohs M, Schleicher R, Schöning J, Essl G, Naumann A, Krüger A (2009) Impact of item density on the utility of visual context in magic lens interactions. Pers Ubiquit Comput 13(8):633-646

43. Roto V, Law E, Vermeeren A, Hoonhout J (eds) (2011) User experience white paper: bringing clarity to the concept of user experience. http://www.allaboutux.org/

44. Salovaara A, Tamminen S (2009) Accept or appropriate? A design-oriented critique on technology acceptance models. In: Saariluoma P, Isomäki H (ed) Future interaction design II. Springer, London, pp 157-173

45. Sato H, Kitahara I, Ohta Y (2009) MR-mirror: a complex of real and virtual mirrors. In: Proceedings of conference on virtual and mixed reality 2009. Springer, Berlin, pp 482-491

46. Satyanarayanan M (2001) Pervasive computing: vision and challenges. IEEE Pers Comm, Aug. 2001, pp 10-17

47. Schmidt A (2002) Ubiquitous computing-computing in context. Dissertation, Lancaster University

48. Stanney K (1995) Realising the full potential of virtual reality: human factors issues that could stand in the way. In: Proceedings of 1995 IEEE virtual reality annual international symposium, IEEE, pp 28−3428-34

49. Swan E, Gabbard J (2005) Survey of user-based experimentation in augmented reality. In: Proceedings of 1st international conference on virtual reality, HCI International 2005

50. Szalavári Z, Eckstein E, Gervautz M (1998) Collaborative gaming in augmented reality. In: Proceedings of the ACM symposium on virtual reality software and technology 1998, pp 195-204

51. Tulving E, Schacter DL, Stark HA (1982) Priming effects in word fragment completion are independent of recognition memory. J Exp Psychol 8(4):336-342

52. Vallino J (1998) Interactive augmented reality. Dissertation, University of Rochester

53. Ware C, Balakrishnan R (1994) Reaching for objects in VR displays: Lag and frame rate. ACM Trans Comp-Hum Interact 1(4):331-356

54. Weiser M (1991) The computer for the 21st century. Sci Am 265(3):94-104

55. Wellner P, Mackay W, Gold R (1993) Back to the real world. Comm ACM 36(7):24-26

56. Wickens CD, Baker P (1995) Cognitive issues in virtual reality. In: Virtual environments and advanced interface design. Oxford University Press, Oxford, pp 516-541

57. Zelkha E, Epstein B, Birrell S, Dodsworth C (1998) From devices to “Ambient Intelligence”. Digital living room conference