A head-based vibrotactile compass for cyclists

Abstract

Cycling stimulates the human mind and body in manifold ways. However, even on a leisure ride, certain waypoints or destinations must be reached. Therefore, orientation is a crucial task for every cyclist. Vibrotactile systems for cyclists do not clog up the visual and auditory senses needed to experience the immediate environment. However, our literature survey shows that previous work focuses on turn-by-turn navigation systems and does not leverage the potential of vibrotactile feedback on the head. Since the head is not in direct contact with the bike, we argue that vibrations occurring naturally are less confounding. Moreover, head movements are already crucial for wayfinding. We present an unobtrusive orientation system for cyclists with head-based vibrotactile feedback – a vibrotactile compass. In a user study, we show the feasibility of our system.

Full text

A head-based vibrotactile compass for cyclists
Anna-Magdalena Krauß
anna-magdalena.krauss@htw-dresden.de
University of Applied Sciences Dresden
Dresden, Saxony, Germany
Dennis Wittchen
dennis.wittchen@htw-dresden.de
University of Applied Sciences Dresden
Dresden, Saxony, Germany
Dietrich Kammer
dietrich.kammer@htw-dresden.de
University of Applied Sciences Dresden
Dresden, Saxony, Germany
Georg Freitag
georg.freitag@htw-dresden.de
University of Applied Sciences Dresden
Dresden, Saxony, Germany
Figure 1: The system supports cyclists in orientation with a headband (right) for vibrotactile feedback, which is activated by a
push button mounted to the bicycle handlebars (left). The accompanying mobile app is not shown in the image.
ABSTRACT
Cycling stimulates the human mind and body in manifold ways.
However, even on a leisure ride, certain waypoints or destinations
must be reached. Therefore, orientation is a crucial task for every
cyclist. Vibrotactile systems for cyclists do not clog up the visual
and auditory senses needed to experience the immediate environ-
ment. However, our literature survey shows that previous work
focuses on turn-by-turn navigation systems and does not leverage
the potential of vibrotactile feedback on the head. Since the head is
not in direct contact with the bike, we argue that vibrations occur-
ring naturally are less confounding. Moreover, head movements
are already crucial for waynding. We present an unobtrusive ori-
entation system for cyclists with head-based vibrotactile feedback
– a vibrotactile compass. In a user study, we show the feasibility of
our system.
Permission to make digital or hard copies of part or all of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for prot or commercial advantage and that copies bear this notice and the full citation
on the rst page. Copyrights for third-party components of this work must be honored.
For all other uses, contact the owner/author(s).
Mensch und Computer 2021, Workshopband , 14. Workshop Be-greifbare Interaktion
©Copyright held by the owner/author(s).
https://doi.org/10.18420/muc2021-mci- ws09-381
CCS CONCEPTS
•Human-centered computing →Haptic devices
;User studies;
Interface design prototyping; User centered design; Activity centered
design.
KEYWORDS
human-computer-interaction, interaction design, head, wearable,
vibrotactile, navigation support, cyclist, user study
1 INTRODUCTION
The purpose of many (digital) tools and processes is to enable peo-
ple to perform their tasks more eciently. In contrast, most leisure
activities are generally free and playful without the need to fulll
specic tasks as fast as possible. In this contribution, we focus on
cycling as such a leisure activity. Although cycling is also used to
reach a certain location, its value is often perceived to be the activ-
ity itself for purposes of recreation, sports, or group experiences.
However, orientation is needed for a cyclist to determine their cur-
rent location in the environment. Most navigation systems follow
a turn-by-turn approach to reach a certain location as eciently as
possible by following a determined path. In contrast, we attempt to
respect the autonomy of the user even more and not distract cyclists
from their environment with unsolicited instructions. Hence, our
goal is to support basic orientation in an unobtrusive way as shown

Mensch und Computer 2021, Workshopband , 14. Workshop Be-greifbare Interaktion Krauß et al.
in our user study. During a bicycle ride, cyclists are exposed to nu-
merous environmental factors that already stress their visual and
auditory senses, such as trac or sightings in nature. Our approach
is based on a vibrotactile interface in order to keep visual and audi-
tory channels free for the actual cycling. The interface is located at
the cyclists’ head, since other parts of the human body are in closer
contact with the bike and could thus be more susceptible to natural
vibrations caused by cycling. The head is also the natural center
for orientation and has a high sensibility for recognizing vibrations
[16].
We present our contributions in this paper as follows: First, we
review related work in the area of vibrotactile interfaces with a
focus on navigation support for cyclists. Second, we describe the
concept for our head-based vibrotactile system. In section 3, we
also outline the technical realization of this system in a working
prototype. This prototype was validated in a user study (section
4). Finally, we discuss the viability of our approach to support
orientation for cyclists (section 4.4) and comment on future work
(section 5).
2 RELATED WORK
Previous research shows how the tactile sense can be leveraged to
convey information via so-called Tactons, generated using vibrotac-
tile displays. Those "structured, abstract messages [...] can be used
to communicate [...] non-visually" [
3
]. In the context of navigation,
they can convey the direction or the distance to the destination. The
system Tacticycle supports tourists in exploring their environment
on a bike [
17
,
18
]. The system combines vibrotactile feedback with
a visual display in a multi-modal way. Potential traveling destina-
tions and points of interests are highlighted on a visual display
mounted on the handlebar. Both handles contain one vibration
motor (tactor) each. Similar to our system, the immediate direction
to the destination or point of interest is conveyed by vibrating.
However, dierent intensities are used and only a 180-degree spec-
trum of directions ahead of the cyclist is covered. The system was
tested in dierent settings following a requirements analysis. In
contrast, Vibrobelt is a turn-by-turn navigation system for urban
areas [
19
]. Eight tactors around the hips direct the cyclist towards
the desired destination. Two distance levels are distinguished by
the duration and repetition of the vibrations. Escobar Alarcón and
Ferrise present vibrotactile wristbands for a navigation system [
1
].
Each wristband contains a single tactor. Consequently, the direc-
tion to the destination is conveyed by vibrations on either the left
or right wrist. Similar to Vibrobelt, two distance levels are distin-
guished by the duration and repetition of the feedback. Matviienko
et al. evaluated dierent modalities (visual, auditory, and tactile)
for navigational cues tailored towards child cyclists [
14
]. Therefore,
they present a navigation system NaviBike similar to Tacticycle,
with a single tactor in each handle of the handlebar. Thus, convey-
ing the direction to the next waypoint and signaling distance in the
same way as with Vibrobelt and Tacticycle. Since hands, feet, and
bottom are in direct contact with the bike they can convey outside
vibrations to the body while cycling on uneven trails. Hence, in
contrast to previous systems, we propose to investigate the head
as location for vibrotactile feedback during cycling. Furthermore,
the head is also the natural center for orientation and therefore we
adapted the hardware setup of Vibrobelt to a headband. Due to the
promising research results of using vibrotactile ring displays on
the head [
4
,
5
,
11
–
13
,
15
,
16
], we aim to leverage this potential to
support orientation for cyclists in a novel, head-based vibrotactile
system.
3 CONCEPT AND IMPLEMENTATION
The metaphor for our approach is a compass, which oers support
for orientation by indicating the direction to north. In our case, a
single reference point serves as the nal destination or an arbitrary
orientation point selected by the cyclist. Moreover, we extended
the metaphor by giving feedback when a destination is reached.
3.1 Requirements
In order to support orientation for cyclists in an unobtrusive and
comfortable way, basic system requirements must be met, following
the standards from ISO 9241 (11, 110, 210) [
8
–
10
]. Simplicity is a
fundamental requirement of the technical system (R1–simplicity).
During the bicycle ride, the outputs should be immediately inter-
pretable and usable for route planning. The frequency of interaction
with the system should be as low as possible (R2–interaction). Fur-
thermore, it should be possible to use the system on longer tours.
An important prerequisite of this is a high level of wearing comfort
in terms of pressure, weight, and vibration (R3–comfort). In addi-
tion, the use of the head-based vibrotactile output system and its
activation should provide high usability (R4–usability).
3.2 Design and Implementation
As shown by systems like Tacticycle [
17
,
18
] and Vibrobelt [
19
],
two-dimensional directional information seems sucient for the
context of cycling. Therefore, we conceived a circular vibrotactile
display, integrated into a headband, that signals the direction to a
destination in relation to the user’s head orientation.
The hardware to realize this system consists of three parts: a
headband, a push button unit, and a smartphone. Since Dobrzynski
et al. showed that twelve tactors are well perceived around the
head [
5
], we followed this suggestion and constructed a vibrotactile
headband with twelve eccentric rotating mass (ERM) tactors. Each
of these are 2
𝑚𝑚
thick and 10
𝑚𝑚
in diameter and were mounted
orthogonally in relation to the head, in order to utilize their vi-
bration direction. This was achieved by inserting them into soft
plastic foam glued onto the headband (see Figure 1 and Figure 2 (2)).
Twelve holes were cut into the outer layer of the headband to route
the tactor cables to the WeMos Lolin32 single-board microcontroller
at the back of the band. The processing unit and a lithium-polymer
battery pack were wrapped by foam rubber and xed by reusable
velcro ties onto the headband. The headband is made of elastic
material, which ts a wide range of head sizes. This preliminary
design for the headband is inferior to a version integrated into an
actual helmet, which is our plan for the future. In total, the head-
band weighs 125 g. To trigger the vibration in a safe manner, we
placed a push button unit consisting of a push button, an indicator
LED, an Arduino HUZZAH32 single-board microcontroller and a
lithium-polymer battery pack on the bicycle’s handlebars. A custom
application running on a smartphone (Google Pixel 2) is used to
set the destination on a map and provides the necessary location

A head-based vibrotactile compass for cyclists Mensch und Computer 2021, Workshopband , 14. Workshop Be-greifbare Interaktion
(4) Android App
communication via
Bluetooth Low Energy
Vibrotactile display
with 12 tactors
(2) Headband
request orientation
information
set destination (once)
and retrieve geolocation
while riding (continously)
(3) Push button(1)
255°
75°
105°
165°
195°
285°
135°225°
315° 45°
15°345°
Figure 2: The vibrotactile compass consists of three main components: the headband as vibrotactile display (2), a push button
to trigger the feedback (3), and a mobile app (4). The headband displays the direction towards the destination with a resolution
of 30 degree (i.e. twelve directions in total) with respect to the user’s head orientation (1).
data while riding. All three hardware components communicate
via Bluetooth Low Energy.
For the directional output, we use simple and concise tactons by
encoding the direction to the destination via the position of a single
activated tactor (see Figure 2 (1)). All tactons used for displaying
directions to the user appear in a repeating sequence of 300
𝑚𝑠
vibration followed by a 300
𝑚𝑠
pause. The duration of the output is
determined by the user, who can request the vibrotactile feedback
in a self-determined manner by pressing the push button (see Figure
2 (3)). The feedback that the destination has been reached must be
clearly distinguishable from the directional tactons described above.
For this purpose, a tacton with a transformation is used. Alternately,
every other tactor is activated for 500
𝑚𝑠
each. To ensure that the
tacton is perceived by the user, it is repeated again after a 500
𝑚𝑠
pause. We purposely omitted other distance information to keep
the system as simple as possible (see R1).
We decided to use a xed vibration frequency for all outputs.
As the manufacturer does not provide a characteristic curve, the
employed frequency can only be estimated by utilizing curves from
similar tactors. By measuring the voltage and using the curve of NFP
310-118, we estimated the vibration frequency to be around 200
𝐻𝑧
and thus slightly higher than recommended by [
4
,
15
] to make
sure that vibrations are perceived impeccably in the conditions of
cycling. Since the frequency and the intensity of the vibration is
coupled for ERMs the intensity remains the same across the tactons
as well.
4 USER STUDY
We validated the requirements dened in section 3.1 for our concept
to test its feasibility for the use case of cycling as a leisure activity.
4.1 Preliminary Test
We tested the output of directions via the headband in a laboratory
setting with 12 participants and in a restricted parking area where
ve given destinations had to be reached using feedback from the
headband. All participants reacted or stated that they perceived the
vibrations. In total, 144 directional outputs were displayed in the
lab and participants aligned themselves to the correct direction 140
times. In 82% of these tasks, the recognition time between presen-
tation of the direction and nal alignment of the participant was
equal or below 10 seconds. In the parking area, 9 of 12 participants
used head rotation for orientation before steering their bicycles
in the appropriate direction. When moving to the target, 11 of 12
participants had to repeatedly correct their route. Due to the lim-
ited dimension of the test eld, circular movements were used to
move closer to the target area. However, 10 of 12 participants rode
directly to the target and only made ne adjustments on the spot.
Hence, we could prove that our headband provides perceivable and
interpretable feedback.
4.2 Study Design
The main study was conducted with six participants (three female,
three male) with an average age of 30,5 (SD=13,05) in a large park
area (1900
𝑚×
950
𝑚
) in the center of Dresden. The objective was
to reach a given destination in a natural navigation scenario. To
make this situation even more realistic, we allowed participants to
use their own bicycles. A common starting point and a destination
unknown to the participants were selected. They were free to nd
their own route without further restrictions (e.g. time limit) and
there was only one test run per participant. An experimenter en-
tered the target into the mobile app before starting the test. After
that, the participant was free to activate the system for orientation

Mensch und Computer 2021, Workshopband , 14. Workshop Be-greifbare Interaktion Krauß et al.
Table 1: Results of the NASA RTLX (0 Very Low - 100 Very
High).
Dimension Mean Median SD Min. Max.
Mental Demand 30 22.5 23.24 5 65
Physical Demand 13.33 17.5 8.76 0 20
Temporal Demand 12.5 15 8.22 0 20
Performance 25.83 22.5 29.40 0 80
Eort 11.67 10 11.25 0 25
Frustration 15 17.5 12.65 0 30
at any time by pressing the push button. Each participant was ac-
companied by two experimenters in approximately 5 to 10 meters
distance in order to observe their general behavior. After the naviga-
tion task, the participants were asked to ll in some questionnaires
to gain qualitative data for the evaluation of R1 (simplicity), R3
(comfort), and R4 (usability).
4.3 Results
First of all, all participants were able to nd and reach the destina-
tion.
Simplicity was investigated using the NASA RTLX test [
6
,
7
].
A total value of 18 (scale 0 - 100, where 0 means easy to use) was
obtained, which means that the system can be used with little
cognitive eort. A detailed list of the results can be found in Table
1. In addition, the participants were asked to recall landmarks or
occurrences seen during their rides (see Figure 3). In particular,
they were requested to state the number of cyclists and families
with children, as well as recalling points of interest (POIs). The
majority of participants were able to indicate the correct number
of cyclists seen (4 of 6) as well as families (5 of 6). Participants who
failed to indicate the correct numbers stated that they had paid
more attention to animals during the ride. All participants were
able to name the correct POIs, even in chronological order. During
the study, participants exhibited relaxed behavior. They were able
to perceive the environment at their leisure and even talk to the
accompanying investigators. The overall results suggest that the
use of the technical setup is pleasant and eortless (R1–simplicity).
In addition, 4 of 6 participants actively looked around by turning
their head. This behavior shows that using our system is in line
with natural head movements of the user for orientation. For better
traceability, the duration and location were saved as soon as the
push button on the handlebar was pressed. The analysis of these
log les showed that the system was only used selectively and
not continuously. The average duration for displaying the signal
was 1
.
12
𝑠
. In addition, all participants used the system ahead of
intersections, while 3 of 6 also used the system on straight stretches
of road to make sure where the destination was located. The study
suggests that the system is mainly used at intersections, but can
also serve as conrmation on straight stretches of track. The rare
interaction with the system met the requirement R2 (interaction).
The requirement for comfort (R3) referred to the wearability
of the headband in general. In a post questionnaire the comfort
was rated as pleasant, with median values of 5 for pressure, 5 for
vibration, and 4.5 for weight (1=unpleasant, 5=pleasant). Overall,
the majority of the participants mentioned that the pressure, vibra-
tion and weight of the headband didn’t bother them at all during
the ride. Furthermore, they stated to not feel restricted in their
movements by the system. In summary, R3 has been met. Usability
(R4) was investigated in the natural setting with a System Usability
Scale (SUS) questionnaire. The average result of 87 points indicates
that the system is “excellent” to “best possible” [
2
]. In the quali-
tative evaluation, the participants also stated that they liked the
system because it aords freedom of movement and ease of use. In
addition, they were able to concentrate completely on the track,
being in control of frequency and duration of the feedback without
being distracted by a navigation display. This consistently positive
feedback shows that R4 (usability) was met in our validation.
4.4 Discussion and Limitations
The aim of the user study was to verify the feasibility of the techni-
cal system. The results showed that the tactons displayed via the
headband are perceptible and distinguishable. With the help of the
tactons, the participants were able to orient themselves in a large
area (approx. 2
𝑘𝑚2
) and to navigate self-determinedly. The head-
band proved to be usable, easy to use, and comfortable. Explicitly,
the high freedom of movement and the immediate interpretation of
the results should be emphasized. In addition, there are individual
dierences and requirements of the cyclist and the chosen route.
Orientation becomes more challenging in unknown environments.
Since most of the participants in the study already knew the test
area, we cannot fully validate the supporting eect of our system
in unknown environments. We need to investigate this in the fu-
ture. The evaluation of comfort showed a high level of satisfaction.
However, the duration of use in the study was shorter than we
intended for a typical bicycle tour that could last more than an
hour. In addition, the headband could only be worn in place of a hel-
met. Integrating the system into a helmet could make a signicant
dierence in safety, perception, and comfort.
5 CONCLUSION AND FUTURE WORK
In this work we presented a vibrotactile feedback system for cyclists
that encodes directional information in order to support orientation.
We conceived and implemented a technical setup that keeps the user
self-determined and conducted a user study to test its feasibility.
The results show that the headband meets our requirements for
simplicity (R1), interaction (R2), comfort (R3), and usability (R4).
For the further development of the system, it is planned to ex-
amine the headband in an in-situ study. This requires investigating
longer and more intensive use during bike tours. Additionally, we
should consider comparing our system to similar (compass-like)
navigation systems. Moreover, we can investigate the level of dis-
traction and cognitive eort of using visual and vibrotactile nav-
igation systems individually or in a combined setup. So far, the
system only provides directional information to the user and there
are various options to extend the corresponding tactons. To in-
crease the degrees of freedom when designing tactons we have to
switch to another type of tactors – namely linear resonant actuator
(LRA). These enable separate control of the vibration’s amplitude

A head-based vibrotactile compass for cyclists Mensch und Computer 2021, Workshopband , 14. Workshop Be-greifbare Interaktion
0100 200 300 m
Figure 3: The routes chosen by the participants in the restricted area.
and frequency. This also allows us to better accommodate the char-
acteristics of vibrotactile perception. Additionally, users can adapt
the vibration’s intensity according to their individual preferences.
From the conceptual perspective, it might be useful to provide more
information such as the distance to the destination. We plan to
publish the schematics and circuit layout as well as the source code
(i.e. rmware and app) under open source licences. This should
encourage others to reproduce the system or even le pull requests
for enhancements or new features.
We are convinced that the suggested setup is not limited to
the context of cycling. Hence, we will consider a transfer to other
helmet systems and user groups, e.g. reghters, visually impaired
people, or industrial workers.
ACKNOWLEDGMENTS
The authors would like to thank all participants for testing the head-
band in our user study. We also thank Philipp Ballin for supporting
the study as experimenter. Especially we thank Alexander Ramian
for his technical support throughout this project. Furthermore, we
thank Christopher Praas for creating the demo video. This work
has been supported by the European Regional Development Fund
and the Free State of Saxony (project no. 741012023).
REFERENCES
[1]
Escobar Alarcon and Francesco Ferrise. 2017. Design of a wearable haptic navi-
gation tool for cyclists. In 2017 International Conference on Innovative Design and
Manufacturing. 1–6.
[2]
Aaron Bangor, Philip Kortum, and James Miller. 2009. Determining what indi-
vidual SUS scores mean: Adding an adjective rating scale. Journal of usability
studies 4, 3 (2009), 114–123.
[3]
Stephen Brewster and Lorna M Brown. 2004. Tactons: structured tactile messages
for non-visual information display. In Australasian User Interface Conference 2004
(ACS Conferences in Research and Practice in Information Technology, Vol. 28).
Australian Computer Society, Sydney, NSW, Australia, 15–23. http://eprints.gla.
ac.uk/3443/
[4]
Victor Adriel de Jesus Oliveira, Luciana Nedel, Anderson Maciel, and Luca Brayda.
2016. Spatial discrimination of vibrotactile stimuli around the head. In 2016 IEEE
Haptics Symposium (HAPTICS) (Philadelphia, PA). IEEE, New York, NY, USA, 1–6.
https://doi.org/10.1109/HAPTICS.2016.7463147
[5]
M. K. Dobrzynski, Seifeddine Mejri, S. Wischmann, and D. Floreano. 2012. Quan-
tifying Information Transfer Through a Head-Attached Vibrotactile Display:
Principles for Design and Control. IEEE Transactions on Biomedical Engineering
59, 7 (2012), 2011–2018. https://doi.org/10.1109/TBME.2012.2196433
[6]
Sandra G. Hart. 2006-10. Nasa-Task Load Index (NASA-TLX); 20 Years Later.
Proceedings of the Human Factors and Ergonomics Society Annual Meeting 50, 9
(2006-10), 904–908. https://doi.org/10.1177/154193120605000909
[7]
Interaction Design Group Magdeburg. 2016. NASA-TLX (Kurzfassung deutsch) -
Beanspruchungshöhe. http://interaction-design- group.de/toolbox/wp-content/
uploads/2016/05/NASA-TLX.pdf Last accessed 14 December 2020.
[8]
ISO 9241-11. 2018. Ergonomics of human-system interaction - Part 11: Usability:
Denitions and concepts.
[9]
ISO 9241-110. 2020. Ergonomics of human-system interaction - Part 110: Interac-
tion principles.
[10]
ISO 9241-210. 2019. Ergonomics of human-system interaction - Part 210: Human-
centred design for interactive systems.
[11]
Oliver Beren Kaul and Michael Rohs. 2016. HapticHead: 3D Guidance and Target
Acquisition through a Vibrotactile Grid. In Proceedings of the 2016 CHI Conference
Extended Abstracts on Human Factors in Computing Systems (San Jose, California,
USA) (CHI EA ’16). Association for Computing Machinery, New York, NY, USA,
2533–2539. https://doi.org/10.1145/2851581.2892355
[12]
Oliver Beren Kaul and Michael Rohs. 2017. HapticHead: A Spherical Vibro-
tactile Grid around the Head for 3D Guidance in Virtual and Augmented Real-
ity. Association for Computing Machinery, New York, NY, USA, 3729–3740.
https://doi.org/10.1145/3025453.3025684
[13]
Oliver Beren Kaul and Michael Rohs. 2019. Concept for Navigating the Visually
Impaired using a Tactile Interface around the Head. In Hacking Blind Navigation
Workshop at CHI ’19 (Glasgow, Scotland Uk) (CHI Workshop ’19). 5 pages. https:
//hci.uni-hannover.de/papers/KaulChi2019_Workshop.pdf
[14]
Andrii Matviienko, Swamy Ananthanarayan, Abdallah El Ali, Wilko Heuten,
and Susanne Boll. 2019. NaviBike: Comparing Unimodal Navigation Cues for
Child Cyclists. Association for Computing Machinery, New York, NY, USA, 1–12.
https://doi.org/10.1145/3290605.3300850
[15]
Kimberly Myles and Joel T. Kalb. 2010. Guidelines for Head Tactile Communication.
Technical Report. Defense Technical Information Center. https://doi.org/10.
21236/ADA519112
[16]
Kimberly P. Myles and Mary S. Binseel. 2009. Exploring the Tactile Modality for
HMDs. In Helmet-mounted displays: sensation, perception, and cognition issues.
U.S. Army Aeromedical Research Laboratory, Alabama, USA, 849–876.
[17]
Martin Pielot, Benjamin Poppinga, Wilko Heuten, and Susanne Boll. 2012. Tacticy-
cle: Supporting Exploratory Bicycle Trips. In Proceedings of the 14th International
Conference on Human-Computer Interaction with Mobile Devices and Services (San
Francisco, California, USA) (MobileHCI ’12). Association for Computing Machin-
ery, New York, NY, USA, 369–378. https://doi.org/10.1145/2371574.2371631
[18]
Benjamin Poppinga, Martin Pielot, and Susanne Boll. 2009. Tacticycle: A Tactile
Display for Supporting Tourists on a Bicycle Trip. In Proceedings of the 11th
International Conference on Human-Computer Interaction with Mobile Devices and
Services (Bonn, Germany) (MobileHCI ’09). Association for Computing Machinery,
New York, NY, USA, Article 41, 4 pages. https://doi.org/10.1145/1613858.1613911
[19]
Haska Steltenpohl and Anders Bouwer. 2013. Vibrobelt: Tactile Navigation Sup-
port for Cyclists. In Proceedings of the 2013 International Conference on Intelligent
User Interfaces (Santa Monica, California, USA) (IUI ’13). Association for Comput-
ing Machinery, New York, NY, USA, 417–426. https://doi.org/10.1145/2449396.
2449450