Keywords

1 Introduction

Virtual reality (VR) technology has been used since the 1960s, but it has been used only in certain areas due to technical problems and high costs. However, it has expanded into the personal domain through a number of factors, including the development of displays and 3D technologies and the miniaturization of hardware, lightening and falling costs [1]. With the recent development of augmented reality (AR) technology, it is being applied in various industries, and the basic technologies are being implemented in different forms [2]. Mobile devices provide users with different interactions and communications from other devices. In general, interaction is divided into person to person interaction and person to machine interaction [3], Mobile devices enable both types of interactions, as well as more and more interaction methods as seen in smartphones. However, there are a lot of wireless application protocol gateways in the study of user interface (UI) reflecting AR, in particular the tendency to focus on navigation. There exists a shortage of professional evaluation of UI design in the AR mobile environment.

With the trend of the times and the change of culture, the requirements of users are multidimensional and varied. The function of a product is different from the way it is used in a service environment, what should be paid attention to is that the rational consumption methods, such as price or function, are changing into the era of emotional service [4]. The perceptual service must respond positively and be applied according to the user’s different request and has the connection with the interface in the process of analyzing and visualizing these requirements [5]. At the same time, an interface that can be close to the users Sensibility should be provided so that the user can experience usability simply and effectively [6], and the USABILITY of software quality, which has been mentioned for a long time, is the core principle of HCI (Human computer Interaction), now it has become a necessary factor [7]. In information appliances such as mobile devices and the Internet, usability is widely used as a means to improve the competitiveness of related products.

This is because usability-considering design can reduce both the physical and mental burdens on users, and improve workers’ productivity and responsiveness [8]. In many cases, more effort is devoted to solving usability problems from the design phase than to post-usability evaluation [9]. It ultimately helps reduce the company’s production and operating costs. Therefore, usability should be evaluated with a combination of different attributes. The ideal usability model is user-focused and friendly, and should easily get users interested in using it and achieving their intended goals [8]. The purpose of designers must be realized together to successfully implement usability. In this research, we examine the usage factors and analyze the linear combination of variable normalization to appraise the usability rapidly. The interface design of AR and the usage of Heuristic approach are studied and reported.

2 Methods

2.1 Heuristic Evaluation

Heuristics.

Heuristics is an abbreviation of the Heuristic evaluation. The heuristic method is for the model solution method, is a method of successive approximation to the optimal solution. In this method, the obtained solution is repeatedly judged and modified until it is satisfied. It refers to the principles that estimators use in evaluating usability. Heuristics varied with the subjects to be evaluated, collated by Nielsen and Molich in 1990 [10].

Heuristic Evaluation.

The heuristic evaluation aims to find the usability problems of the user interface and its advantage that can find problems in the shortest time. Heuristic evaluation method is an expert evaluation method that applies the Heuristics, known as the USABILITY principles, to the interface to produce a usability checklist which was developed by Nielsen and Molich in 1990. It is learned that the Heuristic evaluation method is a relatively low-cost evaluation method that can be applied quickly and easily. Figure 1 is the result of a study conducted by Nielsen and Landauer on the effectiveness of identifying usability problems based on evaluators.

Fig. 1.
figure 1

Proportion of usability problems.

Full size image

The result shows that the efficiency of finding usability problems is about 75% when there are 5 reviewers, but it will decrease sharply when there are more than 5 reviewers. Figure 2 shows the results of a survey on the efficiency of researchers and testing costs.

Fig. 2.
figure 2

Ration of benefits to costs.

Full size image

When there are five reviewers, the ration of benefits of cost was more than 60%. However [11], the ration will reduce with the increase of reviewers. In other words, heuristic evaluation is a method that produces a large number of important usability problems with relatively few expert evaluators, which can execute efficiently and fast.

The evaluation method is to describe the script of the project or to explore the system freely, and the professional estimators are used to judge the accuracy of the prototype design according to the usability principle. However, in an augmented reality environment where the technical environment is not yet mature, the form of interface usability cannot be determined. The existing usable and convenient evaluation methods are most suitable for the development process of software applied in the traditional GUI environments. If they are to be applied to augment reality environments that may not have clear goals, they need to be redefined based on the understanding of the nature of augmented reality.

2.2 Current Stage of Technology and Research

Current Stage of AR Technology.

Augmented reality is a technology that seamlessly blends real-world and virtual world to provide users with a more advanced sense of engagement and reality [12]. Augmented reality has broader implications, especially for the visual basis, display technology, track technology, image integration technology, calibration technology, user interface, interaction, and visualization technology. According to GATTNET’s 2019 trends report, immersive expertise including augmented reality and immersive experiences will change the way we view the digital world, from personal and simple UI technologies to multi-channel and multi-modal experiences. Multi-channel experiences take advantage of not only all human senses but also of advanced computers (heat, humidity, radar, etc.) in multi-mode devices. This multi experience environment provides a rich experience that defines the computer by the space around us [13].

Case Study on Usability Evaluation Elements.

The case study of usability is gathered, centering in Nelson who has compiled principles of the usability evaluation elements. Table  1 describes that the usability studies related to user interfaces are largely based on GUI that leverage existing Heuristic evaluation.

Table 1. Reinterpretation of usability assessment elements
Full size table

The definition of usability depends on the researchers. This paper focus on the definitions of J. Nielsen and R. Mack, J. RoBIN, ISO 1942-11, ISO/IEC 9126, S. Lee, H. Jim, J. KIM.

Nielsen and R. Mack classified the elements of usability into Visibility of system status, Match between system and real world, User control and freedom, Consistency and standards, error prevention, Recognition rather than recall, Flexibility and efficiency of use, Aesthetic and minimalist design, Help users recognize, diagnose, and recover from error, 10. Help and documentation, Skills, Pleasurable and Respectful Interaction with the User, Privacy, 14. Subjective satisfaction, Memorability, Learnability, Efficiency [14].

J. Robin classified it into Core functionality, 1. Should be understandable within an hour, 2. System should be speak the user’s language, 3. System should understand the user’s language, 4. Feedback should be provided for all actions, Feedback should be timely and accurate, 5. UNIX concept should be minimized (ingeneral, minimize underlying system concept), 6. User sensibilities should be considered, 7. Functions should be logically grouped, 8. Interface should be logically ordered, 9. The core function should be clear, 10. The physical interaction whether the system should be natural the system should be efficient, reasonable defaults should be provided, accelerator should be provided, users should not have to enter system – accessible information.

Everything the user needs should be accessible through the GUI (or, in general, through whatever interface style is chosen fir the interface), the user interface should be customized, system should follow real-world conventions, system should follow platform interface conventions, system should be effectively integrated with the rest of the desktop, Keyboard core function should be supported, system should be designed to prevent error, undo and redo should be supported, Good visual design, there is no substitute for a good graphic arts [15]. ISO 1942-11classified it into Effectiveness, Efficiency, Satisfaction, Context of use [16]. While ISO/IEC 9126 classified it into Learnability, Operability, Understandability [17]. S. Lee classified it into Durability, Safety, Size, Familiarity, Arrangement, Attractiveness, Complexity, Simplicity [18].

H. Jim classified it into Effectiveness, Efficiency, Degree of satisfaction, Comprehensibility, Learning ability, Operating ability, Preference, Submissibility, Ease of learning, Empirient performance, System potential. Reuse options, Is it easy to use, Is it easy to learn and teach. Is it easy to learn again, Is it easy to learn or get rid of wrong habits, Is it easy to avoid inconveniences, Is it easy to support, Is it easy to appreciate, Is it easy to share in a group, Is it easy to integrate with traditional methods [19], and J. Kim classified it into Reactivity, Short-cutting, Developmental sensibility, Preventedness, Error repair, Predictability, Familiarity, Generalizability, User liquidity, Controllability, Public property, Multiple-handedness, Personalization, Re-time of change, Comprehensibility [20].

Through the research, the usability evaluation factors can be divided into three kinds. The first is a function-related factor, its purpose is to assess whether the retained functionality is important and whether it meets the user’s requirements. The second is about the evaluation factors of user interaction. Among them, learning, satisfaction, emotional response and attractiveness are all factors. Finally, the third factor is to master the specific issues of design.

2.3 Redefinition of Evaluation Elements

The duplicate items in the survey were removed through the research. A total of 25 usability factors were finally screened out, which are Controllability, Direct Manipulation, Input method, Interaction, User Control, Efficiency, Feedback, Ease, Learnability, Simplicity, Predictability, Consistency, Familiarity, Visibility, Size, Attractiveness, Arrangement, Subjective satisfaction, Effectiveness, Tolerance Principle, Durability, Accuracy, Prevention, Error Indication and Congruity. The selected factors were specified and reinterpreted to suit the augmented reality environment.

According to the usage attributes obtained from the literature survey, the items suitable for augmented reality environment are selected.

For screening, 11 HCI related experts (master’s degree, doctor’s degree) took participated in the questionnaire survey. If the evaluation elements have relevance, they will be assessed with 2 points, if the relevance is not accurate, they will be assessed with 1 point, and if there is no correlation line, they will be assessed as 0 points.

In order to select the usability elements systematically, the statistical method of principal component analysis (PCA) is used.

$$ \begin{aligned} Z & = \left( {\begin{array}{*{20}c} {x_{{1_{1} }} } & \cdots & {x_{{1_{o} }} } \\ \vdots & \ddots & \vdots \\ {x_{{n_{1} }} } & \cdots & {x_{{n_{o} }} } \\ \end{array} } \right)\,n \times p\, \left| { \,F = \left( {\begin{array}{*{20}c} {f_{{1_{1} }} } & \cdots & {f_{{1_{c} }} } \\ \vdots & \ddots & \vdots \\ {f_{{n_{1} }} } & \cdots & {f_{{n_{c} }} } \\ \end{array} } \right)\,n \times c \, } \right| \, \wedge_{c} \\ & = \left( {\begin{array}{*{20}c} {\lambda_{{1_{1} }} } & \cdots & {\lambda_{{1_{c} }} } \\ \vdots & \ddots & \vdots \\ {\lambda_{{p_{1} }} } & \cdots & {\lambda_{{p_{c} }} } \\ \end{array} } \right)\,p \times c\, \left| { \, \Delta = \left( {\begin{array}{*{20}c} {e_{{1_{1} }} } & \cdots & {e_{{1_{c} }} } \\ \vdots & \ddots & \vdots \\ {e_{{n_{1} }} } & \cdots & {e_{{n_{o} }} } \\ \end{array} } \right) \, } \right| \\ \end{aligned} $$
(1)

For the analysis, the following constraints are provided for the four variables Z (standardized observations), F (factor score Matrix), Λ (factor load), and Δ (specific factor), as shown in Eq. 1.

\( \frac{1}{n - 1}\left( {f_{{1_{1} }}^{2} + f_{{2_{1} }}^{2} + \cdots + f_{{n_{1} }}^{2} } \right) = 1 \), The elements are standardized data, with an average of 0 and a dispersion of 1. On this basis, the sum of the squares is equal to n−1.

\( f_{{1_{1} }} f_{{1_{1} }} + f_{{n_{1} }} f_{{n_{1} }} = 0 \), the elements are independence.

\( f_{{1_{1} }} e_{{1_{1} }} + f_{{n_{1} }} e_{{n_{1} }} = 0 \), each factor and independent factor are mutually independent. Moreover, the independent factors are mutually independent. The correlation coefficient and commonness under this constraint are shown by the Eq. 2.

$$ Correlation_{i,j} = \mathop \sum \nolimits_{k = 1}^{c} \lambda_{{i_{1} }} \lambda_{{j_{1} }} = r_{i,j} \left| { Commonity = \mathop \sum \nolimits_{i = 1}^{c} \lambda_{1}^{2} = R^{2} } \right| $$
(2)

Analysis of the results only selected the items with factor loading greater than 0.4. However, Feedback and Interaction consider the interaction to be appropriate and include it as the first factor. In addition, factors such as Learnability, Satisfaction, Uniformity, Consistency consider experience to be appropriate, so they are defined in the second factor. Grasp ability is a physical factor so that making it a third factor. Although it was finally classified as Attractive, it was classified as the fourth element with factor loading higher than 0.4.

The classification factors are shown in Table 2, and the reliability of the four factors is analyzed respectively. We have summarized that

Table 2. Principal component analysis
Full size table
  1. 1.

    Interaction support is the principle that must be provided in the Interaction between user and interface.

  2. 2.

    Experience support is the principle that users feel after Interaction.

  3. 3.

    Learning support is an understanding-related principle in user usage.

  4. 4.

    Emotional support is a principle related to the user’s Emotional support.

  5. 5.

    Cognitive support is the principle that the user should provide in the Cognitive augmented reality UI Model.

The results of reliability analysis were 0.882, 0.931 0.757, 0.827, 0.616. All of them were confirmed to be above 0.6 reliability (Table 3).

Table 3. Reliability analysis of different factors
Full size table

2.4 Discussion and Conclusion

The usability evaluation strategy in this study has reclassified and systematized the current evaluation indexes of GUI objects, focusing on the network and PC in the AR environments. The completed usability taxonomy has been developed as an evaluation system and can be used for both AR usability evaluation and UI design process research in an AR environment. However, up to now, most of the research about AR is technology-focused, but extensive research considering human factors is limited. Further development of the AR environment will require the development of specific evaluation plans that take into account actual users and evaluations of experts and general users.

We have studied and found that specific usability principles for UI in the augmented reality are different from the usability used in the current Interaction support, Learning support Experience, Cognitive support, and Emotional support based on the following principles,

It is essential that all illustrations are clear and legible. Vector graphics (rather than rasterized images) should be used for diagrams and schemas whenever possible. Please check that the lines in line drawings are not interrupted and have a constant width. Grids and details within the figures must be clearly legible and may not be written one on top of the other. Line drawings are to have a resolution of at least 800 dpi (preferably 1200 dpi). The lettering in figures should not use font sizes smaller than 6 pt (~2 mm character height). Figures are to be numbered and to have a caption which should always be positioned under the figures, in contrast to the caption be- longing to a table, which should always appear above the table. Figures and Tables should be cross referred in the text.

  1. 1.

    The element of Interaction support is the element of an application in an existing GUI, but it reflects the Input method feature of the augmented reality. Taking the Input method feature as an example, this study has identified the augmented reality factor for UI effects in a variety of hunting environments as a valid factor, which has not been well studied in depth in the existing GUI environment.

  2. 2.

    Regarding Learning support, simplicity is the specific factor that is reflected in the Learning support factor to prevent confusion from the varied input.

  3. 3.

    In the case of Cognitive support factors, the focus of the specific factor is the importance of natural interactions under augmented reality objectives.

  4. 4.

    If the existing heuristics evaluate the use of fast and simple systems, the concept of sensibility should be emphasized in augmented reality environments.

This work has studied usability elements to make it easy and fast to evaluate usability in the AR interface design. Based on the existing research in the augmented reality environment, the suitable items were extracted with users-focus. 25 items were finally selected with comprehensive statistics. In order to use the final selected usability elements in the augmented reality entity interface, the elements have been redefined. The original concept of the element is materialized to conform to the augmented reality environment. The usability elements were systematized and classified as Interaction support, Learning support, Experience support, Cognition support, Emotion support, etc. Reliability analysis has proved these factors.