=Paper=
{{Paper
|id=Vol-2082/paper_9
|storemode=property
|title=Design Recommendations for HMD-based Assembly Training Tasks
|pdfUrl=https://ceur-ws.org/Vol-2082/paper_9.pdf
|volume=Vol-2082
|authors=Stefan Werrlich,Phuc-Anh Nguyen,Austino-Davis Daniel,Carlos Emilio Franco Yanez,Carolin Lorber,Gunther Notni
|dblpUrl=https://dblp.org/rec/conf/chi/WerrlichNDYLN18
}}
==Design Recommendations for HMD-based Assembly Training Tasks==
https://ceur-ws.org/Vol-2082/paper_9.pdf
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
Design Recommendations for HMD-
based Assembly Training Tasks
Stefan Werrlich Carlos Emilio Franco Yanez Abstract
BMW Group University ITESM In the last few years, head-mounted displays (HMDs)
Stefan.Werrlich@bmw.de cyfranco19@hotmail.com received a growing amount of attention by the scientific
community, especially in the industrial domain. Due to
its possibility to work hands-free while providing the
Phuc-Anh Nguyen Carolin Lorber user with necessary augmented information, HMDs can
BMW Group BMW Group enhance the quality and efficiency of assembly and
Phuc-Anh.Nguyen@bmw.de Carolin.Lorber@bmw.de maintenance tasks. Offering tailored information
requires knowledge about how to design and present
augmented reality (AR) content. However, design
Austino-Davis Daniel Gunther Notni guidelines especially for assembly training tasks as well
BMW Group Technical University Ilmenau as usability evaluations are very limited. In this paper,
Austino-Davis.Daniel@bmw.de Gunther.Notni@tu-ilmenau.de we want to overcome this limitation by introducing an
application as well as 10 design recommendations for
HMD-based assembly training tasks. Furthermore, we
execute a user study with15 participants using an
engine assembly training task to evaluate the software
usability and present results from the system usability
Copyright © 2018 for this paper held by its author(s). Copying permitted scale (SUS) questionnaire, the AttrakDiff as well as the
for private and academic purposes. NASA task load index (NASA-TLX) questionnaire.
Author Keywords
Augmented Reality; Assembly; Evaluation; Head-
Mounted Displays; Training; Usability.
ACM Classification Keywords
H.5.2 [Information interfaces and presentation (e.g.,
HCI)]: User Interfaces
58
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
Introduction section 6. A brief discussion and summary follows at
Augmented Reality (AR) becomes a part of our daily the end of this paper in section 7.
lives. Several applications for smartphones and tablets
are already being used by millions of people. Related Work
Augmented information are designed to improve Design Guidelines are helpful advices for developers
communication, enhance human skills and some of and designers. They provide instructions on how to
them are just for fun. Hand-held devices and projectors adopt specific design principles such as controllability,
are typically used to display superimposed information learnability or customizability. Software design
[1]. In the last years, head-mounted displays (HMDs) recommendations such as the DIN EN ISO 9241-110
received growing interest by researchers in the [5], Shneidermans 8 Golden Rules of Interface Design
industrial domain because they offer a hands-free [6] and the 10 usability heuristics for user interface
usage and help to increase the quality and efficiency of design by Jakob Nielsen [7] are often used for general
assembly and maintenance tasks [2; 3]. In order to software development. Specific guidelines for
design a suitable AR application for manual procedural projection-based AR are presented by Funk [8]. Eight
tasks, researchers have to know the optimal principles, i.e. hands-free usage and personalized
information visualization for different devices. Our feedback, were gained during a four year project using
research is focusing on assembly training tasks because assistive systems for impaired workers. Further specific
they are very important for the automotive industry. recommendations, especially for assembly and
Well executed training must be designed efficiently to maintenance training tasks were published by Webel
ensure a good knowledge transfer whereby optimal [9]. Principles such as mental model building, haptic
process and product quality is guaranteed. However, hints, visual aids and passive learning were introduced
design guidelines for HMD-based applications as well as and focused on acquiring assembly and maintenance
comprehensive usability evaluations are still missing skills which is our focus as well. However, until now it is
[4]. We want to close this gap by providing the still uncertain how to visualize augmented information
following contributions. The second section aims to give efficiently using head-mounted displays (HMDs).
a brief overview of the related work. Our patented Scientific contributions in that field are very rare and
application is introduced and described in section 3. We limited to just a few [10]. We want to overcome this
aim to set this application as the optimal standard for limitation by giving a first suggestion in the next
information visualization using HMDs for assembly chapter.
training tasks. We execute a user study with 15
participants to assess the usability of our application. Application
Detailed information about the experiment are given in This section provides a brief overview of our patented
section 4 and section 5. Due to our gained knowledge application. We describe the relevant functionalities and
during the application development and assessment, show our user interface design. The application was
we extrapolate and present 10 design created using Photoshop for the interface design and
recommendations for HMD-based assembly training in Unity3D for the front-end programming. This
59
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
multimodal application consists of six features with 12]. We added an animated arrow to visualize the end
intuitive icons (Figure 1). of the tunnel. The user can use this augmented tunnel
to find the correct parts in the shelf. This solution
avoids picking mistakes and improves the training
performance due to part search no longer being
required. Another feature provides superimposed
animated 3D information using an outline shader
(Figure 4). The outline visualization is sufficient to
recognize the part’s geometry and position. Additional
Figure 2: Superimposed 3D data arrows show the screw positions. We highly
of an engine part visualized using recommend this visualization technique because it
a head-mounted display. allows to assemble the relevant part without any
superimposition problems. A visualization such as in
Figure 2 may affect the assembly process because the
real part is hard to recognize due to the strong color
rendering.
Figure 1: User Interface for assembly training tasks using
head-mounted displays. 3D feature is activated. The last feature is a video (Figure 5) whereby the user
The trainee can choose between six modalities for each receives detailed information about the current task.
assembly step. A sound feature provides clear auditory Watching a video with an HMD observing someone
instructions about the current task. Another feature performing a task facilitates task transfer. We designed
Figure 3: Augmented Tunnel
visualizes superimposed static 3D data of the the video feature similar to a regular video player. A
Guidance for picking tasks. play and pause button enables the user to have
corresponding part (Figure 2). This feature may help to
learn the position and orientation of the related part. complete control over the feature. The progress bar
When selecting a feature, the icon-color changes to supports the user in building a mental representation of
green and a click sound occurs which gives the user an the task. Additional context information such as a
immediate feedback of his action. Every feature can be progress bar in the middle of the interface as well as
activated and deactivated by either clicking or using the the task overview when selecting the brand icon (Figure
voice command ‘select’. 6) supports mental model building and strengthens the
training transfer. The user receives information about
The text feature provides annotations about the current the finished, the upcoming and the current task. Users
task showing the relevant parts, the activity (e.g. are further able to switch between the assembly steps
assembly or plug), the associated materials such as either by clicking the left or right arrow as well as using
Figure 4: 3D part visualization
using an outline shader and screws and the needed tools. We also implemented a the task overview or using the voice commands ‘next’
arrows to highlight the screw Bezier-curve (Figure 3) which was found to be a good and ‘back’. This concept offers user control and avoids
positions.
solution for picking guidance in previous studies [11;
60
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
a simple step-by-step guidance which is not favorable active and the user is able to continue the assembly
for training tasks. procedure. We used this approach because it offers
several advantages. Backward fading can decrease the
According to the concept of learning introduced by Fitts cognitive workload, enhance the learning transfer and
and Posner [13], we structured our application using improves the initial performance of manual procedural
different learning stages. Information are gradually task [15].
reduced during the training. All features are available in
the tutorial level. The first level is made for exploring Evaluation
the task, the application as well as familiarizing with We conducted a user study with 15 participants to
Figure 5: Augmented Reality
video player for assembly training using a HMD. Two features with the strongest guidance evaluate the usability of our HMD-based software for
tasks. were blocked in the beginner level. The augmented assembly training tasks. This section describes the
tunnel and the outline features were equipped with a study design, explains the procedure, introduces the
bolt sign to visualize the restriction (Figure 7). When hardware setup, gives a detailed information about the
clicking on one of the restricted feature, the avatar (we participants and reports the results of our
named Embly) loses one of its seven lives (Figure 8). measurements.
This game-based learning approach aims to motivate
the user finishing the task autonomous using the Design
available features without killing Embly. Two more To evaluate the usability of our training software for
functions, the 3D feature as well as the video function assembly training tasks, we designed an experiment
Figure 6: Assembly task are additionally blocked in the intermediate level. Every with three groups and different knowledge backgrounds
overview.
feature is restricted in the expert level. Only a default in AR and assembly processes (independent variables)
audio with information about the underlying task is following a between-subject design recommended by
provided for each step. Nielsen [16]. Measuring the usability of a software
includes the assessment of effectivity, efficiency and
Additionally, we used a single backward fading learning user satisfaction variables [17]. To gather the
approach which was found to be effective for learning effectivity of our training software we measured the
by Renkl [14]. This means, the last step is faded out in dependent variables assembly (AM) and picking
the tutorial level, the last two steps in the beginner mistakes (PM), self-corrected assembly (CAM) and
level, the last three in the intermediate level and the picking mistakes (CPM) as well as correction by help for
last four steps in the expert level. The user is asked to the assembly (CBHA) and picking (CBHP). We further
select the correct part before receiving information verify the backward fading questions (BWF) allocating
Figure 7: Restricted User
Interface in the Beginner Level. about the task (Figure 9). At this time, all six features either one point (correct answer) or zero points (wrong
are blocked. Participants receive a visual (green color) answer). As dependent variable for the efficiency, we
and an auditory feedback as soon as they select the measured the tutorial level completion time (TLCT), the
right part. The part is marked red in color if the user beginner level completion time (BLCT), the
selects the wrong part. Afterwards, all features become intermediate level completion time (ILCT) as well as the
61
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
expert level completion time (ELCT). Additionally, we Procedure
collect user satisfaction data using the extended system Through a public invitation, we acquired five office
usability score (SUS) according to Bangor [18] as well employees, five assembly employees as well as five AR-
as the AttrakDiff questionnaire [19]. We also used the experts in preparation of our study. All participants
NASA Task Load Index (NASA-TLX) to rate the were informed in advance to bring safety boots and
perceived workload during the experiment [20]. We safety gloves. We initially made all participants familiar
measured six dependent variables, the mental workload with the environment since the test environment and
(MWL), the physical workload (PWL), the temporal the assembly task was new for every participant. At
workload (TWL), the user performance (UP), the user first, we explained the assistive system and informed
Figure 8: User activated a effort (UE) as well as the user frustration (UF). A high every participant that their participation is voluntary.
restricted feature wherby Embly cognitive workload may harm the learning process We further told them to inform us whenever they feel
loses one of its seven lives.
because fewer cognitive resources are available which uncomfortable so we can abort the experiment
are needed to store relevant information in the immediately. Afterwards, we explained the purpose of
procedural memory. the usability study. After explaining the ambition of the
experiment, we measured and adjusted the user’s
Apparatus interpupillary distance (IPD) which is important for the
For our experiment, we used a Microsoft HoloLens HMD visual quality. Holograms may appear unstable or at an
to display our training software, providing all assembly incorrect distance when using an incorrect IPD. We
instructions. In contrast to other researchers who used showed how to adjust the HoloLens and started with
Figure 9: Embly asked the user
low complex Lego Duplo assembly task to evaluate the Microsoft Learn Gestures Application to familiarize
to select the correct part using
backward fading. their solutions [21], we used a real engine assembly our participants with the interaction modalities. Once
task. The test environment was build referring to the the participants felt confident using the HMD, we kindly
production workplace (Figure 10). The workplace asked our users to start our training application. All
consists of three areas. A shelf area providing all the participants were asked to complete the tutorial level at
parts and screws necessary for the assembly process. first, continuing with the beginner, intermediate and
All tools can be found in the tool area. The assembly expert levels. Users had to work through 15 assembly
area includes a driverless transport system (DTS) steps in each level. The assembly sequence between
mounted with a six-cylinder engine. We used an the levels was not modified. Only the provided
assembly training task with 15 steps following information were reduced. Between each level, we
production specification. The training contains low disassembled the engine back to its initial state. During
complexity tasks such as screwing a lifting eyebolt but that time, participants were asked to have a 10
also high complexity tasks such as installing, screwing minutes break. We offered various sweets and soft-
Figure 10: Work environment for and plugging a harness. drinks to generate a pleasant break. During the study
the engine assembly training. we measured the time for each level, assembly and
picking mistakes, self-corrected mistakes, corrections
by help as well as the backward fading questions. To
62
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
measure the training time between each assembly we SD = 0.00) and medium experience with assembly
paired our application with a database using a WiFi processes (M= 2.00; SD = 0.63). All participants were
internet connection. This approach guarantees a capable to understand, read and write the German
reliable data collection. After finishing the fourth and language since the entire auditory instructions provided
last level, participants were asked to rate the training by our software as well as the questionnaires were in
software using three established questionnaires. We German.
used an extended SUS to evaluate the usability of our
software. Participants had to finish a 10 item Results
questionnaire using a five options Likert scale ranging There was no significant difference between the
from strongly agree to strongly disagree. The second assembly training times (Table 3). The Shapiro Wilk
Table 1: Assembly mistakes questionnaire (AttrakDiff) aims to determine the Test showed a normal distribution for TLCT, ILCT and
during the assembly training
pragmatic and hedonic quality. The questionnaire was ELCT and non normal distribution for BLCT (p=.02). We
using HMD-based instructions.
finished with the NASA-TLX to assess the cognitive used a one way ANOVA which showed no a statistically
workload during the training. significant difference for the TLCT (F(2,12)=.478;
p=.631) and ELCT (F (2,12)=1,189; p=.388). The ILCT
Participants did violate the variance homogeneity (p=.034).
We invited 15 participants (13 male, 2 female) for our Therefore we used the Welch Test which showed no
user study following Nielsen who recommends significant difference for the ILCT (F(2, 7,425)=1,855;
performing a usability study using three groups with p=.222). The Kruskall Wallis Test for BLCT also showed
five users each [16]. The participants were aged from no significant difference (χ² (2)=.08 ; p=.961) between
21 to 42 (M = 30.06; SD = 6.20). Five of them were the groups.
Table 2: Picking mistakes during
office employees, five were assembly employees
the assembly training using HMD-
based instructions. working in the BMW Group production and five were During the study, all three groups made a few errors
AR-experts with at least 5 years background in AR. We (Table 1; 2) but there was so significant difference
asked each group for their AR and assembly between the groups. The Shapiro Wilk Test did show a
background using a five item Likert scale ranging from non-normal distribution for all variables (AM, PM, CAM,
Table 3: Average level
much experience to few experience. Much experience CPM, CBHA, CBHP). We used the Kruskall Wallis Test to
completion times in seconds.
were scored with 4 points, few experience with 0 find difference between the variables but the test
points. Office employees stated to have no background showed no significant difference for AM (χ² (2)=1,227 ;
in AR (M= 0.20; SD = 0.40) and medium experience p=.541), for PM (χ² (2)=1,745 ; p=.418), for CAM (χ²
with assembly processes (M= 1.60; SD = 1.35). The (2)=.162 ; p=.922), for CPM (χ² (2)=1,536 ; p=.458),
assembly workers had a strong background in assembly for CBHA (χ² (2)=.560 ; p=.756) and for CBHP (χ²
processes since it’s their daily routine (M = 3.80; SD= (2)=.126 ; p=.939) between the groups. We also found
0.40) but their knowledge about AR was limited (M = no significant difference for BWF (χ² (2)=3,960 ;
0.40; SD = 0.49). In contrast to that, all AR-experts p=.138).
stated to have a strong background in AR (M = 4.00;
63
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
We used the NASA-TLX to measure the mental Design Simple.
workload during the experiment but there was no We highly recommend to use a simple, clear
significant difference for the six subscales between the understandable, consistent application design. Low
groups. The Shapiro Wilk Test did show a non-normal complexity designs and uniform colors help to reduce
distribution for PW (p=.029) and TWL (p=.038). The the cognitive workload which improves the training
Kruskall Wallis Test showed no significant difference for transfer. However, visual complexity increases the
PW (χ²(2)=1,831; p=.40) and TWL (χ²(2)=2,964; brain activity and therefore the cognitive workload
p=.227) between the groups. The variables MWL, UP, which harms the procedural memory.
UE, UF showed a normal distribution and the one way
ANOVA showed no a statistically significant Enable users to control the software.
difference for MWL (F(2,12)=2,406; p=.132), for UP Guiding a user step-by-step through a procedural task
(F (2,12)=.421; p=.666), UE (F (2,12)=2,983; improves the initial performance but may have an
Figure 11: AttrakkDiff result with p=.089), UF (F (2,12)=.097; p=.908). adverse effect on the training transfer. Users might not
a pragmatic quality of 1.63 and a be able to repeat a task without having the support
hedonic quality of 1.40 for n=15.
The results from the SUS also showed no significant from an assistive system. Instead of just guiding a user
difference (F(2,12)=.425; p=.663) between the groups. through a task, they should have full control of the
Nine participants stated that the application was application. Users are able to activate and deactivate
‘excellent’, six of them said it was ‘good’. The SUS different features, switching between next and previous
showed an average score of 90.5 (M= 90.5; SD= 4.76) assembly steps as well as different levels of difficulty at
which indicates a high user satisfaction. Additionally, any time, when using our software. This allows users to
we asked all participants to rate the hedonic (HQ) and work self-paced without feeling like a robot.
pragmatic quality (PQ) of our software using the
AttrakDiff questionnaire. Results indicate a PQ of 1.63 Provide multimodal feedback.
with a confidence of 0.25 and a HQ of 1.40 with a Humans are used to multimodal feedback since it is
confidence of 0.38. Users desire this application for provided by a lot of technical systems in our daily lives
future use in assembly trainings (Figure 11). (e.g. visual and sound feedback is provided when
pressing a button in an elevator). Adapting familiar
Recommendations feedback approaches to new technologies such as
Based on the experiences we gained from using our HMDs helps to familiarize new users in a shorter
HMD-based application with different user groups, we amount of time. Furthermore, the combination of
propose 10 recommendations for designing HMD-based different modalities such as visual, auditory or tactile
assembly training tasks. We also believe that these feedback can improve the learning transfer through
guidelines are generic and easily transferable to other stimulating various human information channels.
procedural tasks and different domains (e.g. learning a
surgery procedure) which could help designers and
researchers to create more meaningful applications.
64
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
Offer different user modes. Add context information.
Providing a wide range of information at the beginning Adding context information can help users to build and
of a learning process helps novice users gather strengthen a global picture of a task. Having a strong
essential information of the task. Our evaluation mental model of a procedure allows someone to
revealed that once users completed the tutorial level, perform a task efficiently without requiring support
they become much more familiar in using the software from an assistive system (e.g. HMD). Due to our user
modalities as well as performing basic movements study, we recommend using progress bars for
which indicates the time improvement in Table 3. At visualizing step-by-step progresses as well as when
this point, information should be gradually reduced by implementing video players to present the length of a
not frustrating users with too much unnecessary video. A complete task overview was also found to be
information. We recommend offering different user helpful by our participants for building a global picture
modes ranging from a tutorial to an expert mode using of the assembly training task.
various amounts of information. This concept allows
completely novice employees as well as experienced Integrate gamification elements.
users to use a software product in accordance to their Assembly processes are often boring and monotonous
skill level. for many employees. When it comes to learning new
procedures, the majority are willing to acquire new
User voice interaction. knowledge but they wish more fun during the training
Executing manual procedural task requires hands-free since it takes a lot of time and is very serious. Game-
usage which is usually realized when using a HMD. based learning approaches which are already
Additionally, developers should take into account that widespread in schools for teaching kids might help to
extra effort as well as limitations in performing a task improve assembly trainings. Previous studies already
should be avoided when designing interaction concepts. stated that providing self-quantified information such
Therefore, we suggest using voice interaction due to as errors and time, combined with gamification
two facts. Most of the time, users carry parts and tools elements can enhance work processes [23]. Our
when performing assembly tasks. They should be able participants revealed that Embly is cute and motivates
to interact with the HMD without using their hands them to finish the assembly training successfully.
which can be realized using voice interaction. Through Designing our application according to a game was also
our study, we also learned that gesture interaction, found to be very enjoyable by our participants. We also
especially the HoloLens Airtap was hard for many believe that the attractiveness of assembly jobs can be
participants since it’s an unnatural movement. More increased for younger people when providing innovative
natural and intuitive gesture interaction concepts may technologies such as HMDs in combination with game-
help to overcome this limitation in future [22]. based learning applications.
65
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
Present multimodal information. when presenting animated 3D parts. While watching a
When analyzing the click rates of our application, we looped animation of the assembly process, users are
found differences between users and specific user able to install the part simultaneously. Rendering only
Table 4: Feature click rates
during the assembly training. groups (Table 4). Most of our participants tend to use full-color shader parts while wearing a HMD might
traditional, familiar media such as watching a video or bother users executing the assembly process because
reading a text. On the other side, some participants real parts are difficult to recognize. Users might tend to
also preferred using three-dimensional content. watch underneath the HMD to accomplish the assembly
Therefore, providing the opportunity to choose between process.
different types of information will help many users
finish an assembly task successfully. Additionally, Discussion and Conclusion
offering multimodal information can enhance the In this paper, we introduced a novel concept for
training transfer. assembly training task using HMDs. According to the
requirements for industrial killer applications introduced
Build a clean multilayered architecture. by Navab [24], we build a reliable, scalable, user
People are familiar working with multilayered mobile friendly killer application for real engine assembly
phone applications. Each layer contains different training tasks and described every feature in detail. An
information and functionalities. Taking this into account experiment with 15 participants, divided into three
when developing applications for HMDs can help novice groups with different skill levels, were executed to
users to become adjusted to a new software and evaluate the usability of the software. Results regarding
technology very fast. We adapted this approach and effectivity, efficiency and user satisfaction variables
designed a clean multilayered software for assembly showed no significant differences between the three
training tasks. Only the UI (Figure 1) is visualized groups. One reason for that might be the low number
permanently, all other functionalities and information of participants. Therefore, we argue that everyone can
are hidden under sublayers, selectable on demand. We use this application, no matter which skill level or
recommend using this concept since all of our knowledge background someone has. Due to that fact,
participants liked it. we have created a standard tool for assembly training
task which ensures a consistent educational result. All
Visualize different 3D content. participant enjoyed using the HMD-based application
Superimposed three-dimensional content supports proven by the high SUS of 90.5 and the result from the
trainees in learning the position and orientation of a AttrakkDiff. Based on the experiences during the
specific part as well as improves the spatial perception. evaluation, we proposed 10 design recommendations
Providing additional animations not only helps to for HMD-based assembly training. These principles can
understand what and where to assemble, it also shows be adopted by other researches to create a successful
how to assemble a specific part. Therefore, we training application. Considering future work, we want
recommend using different 3D augmented reality to evaluate our application with a larger amount of
content. We further suggest using an outline shader participants by measuring the trainer transfer.
66
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
Acknowledgements [6] Shneiderman, Ben (1998): Designing the user
We thank all participants for their unaffordable input and interface: Strategies for effective human-
BMW, for helping and providing with the tools, spaces computer interaction. ISBN: 0201694972.
and funding to realize and complete this project
successfully. [7] Nielsen, Jakob; Molich, Rolf (1990): Heuristic
evaluation of user interfaces. In: Chew, Whiteside
References (Hg.) 1990 – Proceedings of the SIGCHI
[1] Büttner, Sebastian; Mucha, Henrik; Funk, Markus; conference, S. 249–256. DOI:
Kosch, Thomas; Aehnelt, Mario; Robert, 10.1145/97243.97281.
Sebastian; Röcker, Carsten (2017): The Design
[8] Funk, Markus; Kosch, Thomas; Kettner, Romina;
Space of Augmented and Virtual Reality
Korn, Oliver; Schmidt, Albert (2016): motionEAP:
Applications for Assistive Environments in
An Overview of 4 Years of Combining Industrial
Manufacturing. In: Proceedings of the 10th
Assembly with Augmented Reality for Industry
International Conference on PErvasive
4.0. In: Proceedings of the 16th international
Technologies Related to Assistive Environments,
conference on knowledge technologies and data-
S. 433–440. DOI: 10.1145/3056540.3076193.
driven business.
[2] Hou, Lei; Wang, Xiangyu; Bernold, Leonhard;
[9] Webel, Sabine; Bockholt, Ulrich; Engelke, Timo;
Love, Peter E. D. (2013): Using Animated
Gavish, Nirit; Tecchia, Franco (2011): Design
Augmented Reality to Cognitively Guide Assembly.
Recommendations for Augmented Reality based
In: Journal of Computing in Civil Engineering, S.
Training of Maintenance Skills. In: Recent Trends
439–451. DOI: 10.1061/(ASCE)CP.1943-
of Mobile Collaborative Augmented Reality
5487.0000184.
Systems, S. 69–82. DOI: 10.1007/978-1-4419-
[3] Hořejší, Petr (2015): Augmented Reality System 9845-3_5.
for Virtual Training of Parts Assembly. In: Procedia
[10] Werrlich, Stefan; Eichstetter, Elisabeth; Nitsche,
Engineering 100, S. 699–706. DOI:
Kai; Notni, Gunther (2017): An Overview of
10.1016/j.proeng.2015.01.422.
Evaluations Using Augmented Reality for Assembly
[4] Wang, X.; Ong, S. K.; Nee, A. Y. C. (2016): A Training Tasks. In: International Journal of
comprehensive survey of augmented reality Computer and Information Engineering, S. 1074–
assembly research. In: Advances in 1080.
Manufacturing, S. 1–22. DOI: 10.1007/s40436-
[11] Bjoern Schwerdtfeger (2012): Pick-by-vision:
015-0131-4.
Bringing HMD-based Augmented Reality into the
[5] Ergonomics of human-system interaction - Part Warehouse. ISBN: 978-3832526276.
110: Dialogue principles (ISO 9241-110:2006);
[12] Biocca, F.; Tang, A.; Owen, C.; Fan, Xiao (2006):
Deutsche Fassung EN ISO 9241-110:2006.
The Omnidirectional Attention Funnel: A Dynamic
67
AR in the Industry SmartObjects '18, in conjunction with CHI '18, Montreal, Canada
3D Cursor for Mobile Augmented Reality Systems. [19] Hassenzahl, Marc; Burmester, Michael; Koller,
In: Proceedings of the 39th Annual Hawaii Franz (2003): AttrakDiff: Ein Fragebogen zur
International Conference on System Sciences Messung wahrgenommener hedonischer und
(HICSS'06), S. 22c-22c. DOI: pragmatischer Qualität. In: Szwillus (Hg.) 2003 -
10.1109/HICSS.2006.476. Mensch & Computer 2003, S. 187–196. DOI:
10.1007/978-3-322-80058-9_19.
[13] Fitts, P.M; Posner, M.I (1967): Human
Performance. In: Belomnt, CA: Brooks/Cole. [20] Hart, Sandra G.; Staveland, Lowell E. ():
Development of NASA-TLX (Task Load Index):
[14] Renkl, A.; Atkinson, R.K; Maier, U.H; Staley, R.
Results of Empirical and Theoretical Research. In:
(2002): From example study to problem solving:
Meshkati, Hancock (Hg.) 1988 - Human Mental
Smooth transitions help learning. In: Journal of
Workload, S. 139–183. DOI: 10.1016/S0166-
Experimental Education, S. 293–315.
4115(08)62386-9.
[15] Eiriksdottir, E.; Catrambone, R. (2011):
[21] Loch, Frieder; Quint, Fabian; Brishtel, Iuliia
Procedural Instructions, Principles, and Examples:
(2016): Comparing Video and Augmented Reality
How to Structure Instructions for Procedural Tasks
Assistance in Manual Assembly. In: 2016 12th
to Enhance Performance, Learning, and Transfer.
International Conference, S. 147–150. DOI:
In: Human Factors: The Journal of the Human
10.1109/IE.2016.31.
Factors and Ergonomics Society, S. 749–770.
DOI: 10.1177/0018720811419154. [22] Funk, Markus; Kritzler, Mareike; Michahelles,
Florian (2017): HoloLens is more than air Tap. In:
[16] Nielsen, Jakob; Landauer, Thomas K. (1993): A
IoT '17 Proceedings of the Seventh International
mathematical model of the finding of usability
Conference on the Internet of Things, S. 1–2.
problems. In: Arnold, van der Veer et al. (Hg.)
DOI: 10.1145/3131542.3140267.
1993 - Proceedings of the SIGCHI conference, S.
206–213. DOI: 10.1145/169059.169166. [23] Korn, Oliver; Muschik, Peter; Schmidt, Albrecht
(2016): Gamification of Production? A Study on
[17] Ergonomische Anforderungen für Bürotätigkeiten
the Acceptance of Gamified Work Processes in the
mit Bildschirmgeräten, Teil 11: Anforderungen an
Automotive Industry. In: Advances in Affective
die Gebrauchstauglichkeit - Leitsätze; Deutsche
and Pleasurable Design, S. 433–445. DOI:
Fassung EN ISO 9241-11:1998.
10.1007/978-3-319-41661-8_42.
[18] Bangor, Aaron; Kortum, Philip T.; Miller, James T.
[24] Navab, Nassir (2004): Developing killer apps for
(2008): An Empirical Evaluation of the System
industrial augmented reality. In: IEEE Computer
Usability Scale. In: International Journal of
Graphics and Applications, S. 16–20. DOI:
Human-Computer Interaction, S. 574–594. DOI:
10.1109/MCG.2004.1297006.
10.1080/10447310802205776.
68