Abstract— Flatland was a large scale immersive theatre
production completed in March 2015 that made use of a novel
shape-changing haptic navigation device, the ‘Animotus’.
Copies of this device were given to each audience member in
order to guide them through a 112m
dark space to large tactile
structures accompanied by audio narration from the
production’s plot. The Animotus was designed to provide
unobtrusive navigation feedback over extended periods of time,
via modification of its natural cube shape to simultaneously
indicate proximity and heading information to navigational
targets. Prepared by an interdisciplinary team of blind and
sighted specialists, Flatland is part performance, part in-the-
wild user study. Such an environment presents a unique
opportunity for testing new forms of technology and theatre
concepts with large numbers of participants (94 in this case).
The artistic aims of the project were to use sensory substitution
facilitated exploration to investigate comparable cultural
experiences for blind and sighted attendees. Technical goals
were to experiment with novel haptic navigational concepts,
which may be applied to various other scenarios, including
typical outdoor pedestrian navigation. This short paper
outlines the project aims, haptic technology design motivation
and initial evaluation of resulting audience navigational ability
and qualitative reactions to the Animotus.
Comparable experiences for visually impaired (VI) and
sighted individuals are rarely achieved in daily life. Often ‘VI
accessible’ versions of cultural experiences (e.g. in
entertainment or the arts) provide large amounts of visual
stimulus to people who can see, but limited audio
descriptions to those who are VI. Examples include movies
with additional audio descriptions. This creates a
discrepancy, where the medium on display has been designed
for sighted persons and later retrofitted for a VI minority.
In our work we seek to explore the possibility of
designing immersive promenade theatre experiences for both
sighted and VI groups. To achieve this aim we make use of a
pitch black environment and haptic sensory augmentation
technology. This aims to level the sensory abilities of both
groups (VI and sighted) as they are placed together into an
unfamiliar space, the exploration of which is encouraged via
the theatrical setting. Promenade / immersive theatre is
defined as theatre where the audience can move about to
explore the piece rather than remaining stationary.
*Research supported by NESTA Digital Fund, London, UK.
A. J. Spiers and A. M. Dollar are with the Department of Mechanical
Engineering, Yale University, New Haven, CT 06511, USA. (phone: 203-
432-4380, e-mail: email@example.com, firstname.lastname@example.org).
J. van der Linden and S. Wiseman are with the Pervasive Interaction
Lab, Open University, Milton Keynes, UK (j.vanderLinden@open.ac.uk)
M. Oshodi is with the Extant theatre company, London, UK
As in most theatre, Flatland features a plot and characters.
These were adapted from the 19
century novella Flatland,
by E. Abbot . An initial portion of this plot is explained by
an actor during an introductory session. Subsequent elements
of the plot may then be uncovered by locating zones in the
performance space, each of which is defined by a large tactile
set piece and audio narrative (delivered through wireless
bone conducting headphones). In order to locate these set
pieces and uncover the story, a haptic sensory substitution
device, the Animotus (Figure 1), is provided to each audience
member. This device was designed with the intention of
presenting highly intuitive navigation assistance without
distracting from the overall theater experience. This led to the
choice of haptic shape changing feedback as the interface for
simultaneously communicating both heading and proximity
to the next zone, with continuous 100Hz updates. Intuitive
and unobtrusive mutli-DOF haptic feedback is believed to be
useful outside of this specific application, to enable a discrete
alternatives to screen and audio based pedestrian navigation
in real world scenarios, for both sighted and VI persons.
Though this project is limited to a specific indoor
environment, the navigation technology was designed with
consideration of real world application to unstructured
spaces, such as typical outdoor (and indoor) pedestrian
navigation scenarios, complete with the constraints of
sidewalks, corridors and obstacles. The technology was also
designed to be ‘inclusive’, to benefit both VI and sighted
individuals when the environment is not necessarily dark.
Though the advent of GPS and smartphones have made
navigation guidance while walking outdoors commonplace,
the main interface for this technology is screen based. In 
this was considered (for sighted persons) as potentially
distracting from various hazards and a possible cause of
increasing mobile phone related accidents . While screens
are inaccessible for severely VI persons, the use of audio
instructions during GPS navigation is used by many.
Flatland: An Immersive Theatre Experience Centered on
Shape Changing Haptic Navigation Technology
Adam J. Spiers Member, IEEE, Janet van Der Linden,
Maria Oshodi, Sarah Wiseman, Aaron M. Dollar, Senior Member, IEEE
Figure 1: (Left) Audience members equipped with localization
equipment, bone conducting headphones and Animotus (right)
However, the requirement of headphones in noisy urban
environments can mask the ambient sounds used to avoid
hazards, appreciate one’s surroundings or communicate with
others . Haptic interfaces may provide a more appropriate
stimulus to both VI and sighted groups, due to the less critical
role of touch during walking. Indeed, the most successful and
long-standing VI mobility aids are the guide cane and guide
dog, which both provide feedback by haptic cues delivered
through the cane’s handle or dog’s harness. The appeal and
benefit of haptic navigation to sighted individuals is also
apparent in widespread consumer interest in the ‘Taptic’
interface of the recent Apple watch, which is capable to
provide haptic navigation instructions .
Various haptic navigation and motion guidance systems
have been proposed, often for reasons similar to those
described above. The potential of haptics to provide sensory
augmentation without drawing on critical attentional
resources (i.e. sound and, if applicable, sight) has great
appeal . Though many haptic sensations exist, the authors
of  highlighted that a frequent choice for motion guidance
applications has been vibrotactile feedback (e.g. ).
This technology has many benefits (the actuators are small,
lightweight, inexpensive, low power and easy to control).
Vibrotactile feedback is now standard feature in mobile
phones where it is primarily used to signify discrete and
(generally) infrequent events, such as a new message or
incoming call. In  the success of such feedback is
attributed to the ‘firm fit with the usability constraints of
signifying alerts’. Other authors  have suggested that
alerts are not always an appropriate form of information
delivery and that designers of technology should consider a
haptic stimuli’s place in a user’s attention spectrum, so that it
does not distract from more critical tasks. In our work, the
goal was to present users with frequent navigation guidance
over potentially long periods of time (up to 50 minutes),
without interfering with the user’s appreciation of the
Flatland theatre experience. As such, ‘alerting’ stimuli were
deliberately avoided, as it was felt that frequent high-
attention feedback over such a time scale may become
distracting or tiresome, as also observed in –.
In  and  a number of wearable or chair-mounted tactile
feedback devices are proposed that aimed to avoid ‘alerting’
sensations of other feedback modes. In , Hemmert et al.
proposed the use of shape changing handheld objects to
indicate direction in a simulated navigational task (users
matched the indicated direction by turning an office chair).
Considering such modalities as inspiration, the Animotus
(Figure 1) was designed as a handheld haptic device that
could provide constant navigational guidance over extended
periods of time by changing shape in the user’s hand.
The Animotus was developed through multiple diverse
prototypes, focus sessions with sighted and visually impaired
members of the Flatland production team and laboratory
based user testing (the outcomes of which will be reported on
in separate papers). The final version of the device and its
associated articulation is illustrated in Figure 2. When in its
home pose (Figure 2a), the device shape approximates a cube
with rounded edges and dimensions (60×60×40mm). The top
half of this cube is able to independently rotate (±30deg) and
translate (11.75mm) relative to the bottom half. This allows
the device to indicate direction (heading) and distance
(proximity) to a navigational target. An embossed triangle on
the top of the Animotus is a simple tactile feature to allow a
user to identify the top and front of the device, when it cannot
be seen. A tactile groove that traverses the front face aids
heading perception by aligning when heading error is 0deg.
By holding the device in their upturned (supinated) hand
(Figure 1) a user naturally wraps their fingers around the
front and sides of the device in a power grasp. The height of
the device was selected to permit this grasp to be achieved
for a variety of hand sizes. In this grasp the bottom half of the
device is grounded on the user’s palm, while the relative pose
of the top half may be felt by the user’s fingers. The force
and torque exertion capability of the linear and rotational
DOF are 25N and 1Nm respectively, allowing the device to
exert sufficient forces to achieve motion, even when gripped
tightly. The device weighs 105g and is 3D printed in ABS.
Eight of these devices were built, at a cost of $75 each. Each
Animotus is controlled by an X-OSC wireless
microcontroller and powered by a LiPo battery (120g
combined weight). These are worn by the user in a pouch and
connected to the Animotus via cables, though future
iterations may integrate the components into the device.
Within the Flatland environment (16 x 7 meters), each
Animotus served to direct its audience member from one
zone to the next, allowing them to gradually uncover the
production’s plot. An illustration of the environment is
presented in Figure 3. Note that most zones have separate
exits and entrances, though not all audience members
adhered to these. Each Animotus responds to the position and
orientation of its user (audience member) by continuously
updating its extension and rotation axes (at 100Hz), with
respect to the current navigational target. Audience position
was measured via a Ubisense localization system, via small
active radio tags (weight 40g) worn by each audience
member. Orientation was measured via a wrist worn, tilt
compensated magnetometer. Together these systems allowed
wireless localization of individuals with 0.4m / 2deg
accuracy at 100Hz. A centralized navigation computer
compared user position and orientation with the co-ordinates
Figure 2: The Animotus a) in home pose, b) illustrating rotation
and linear extension.
These DOF may
Figure 3: The performance environment, showing (E) zone
entrances, (X) exits, (F) final exit and physical zone structures.
of the virtual navigational targets (the entrances to the
zones). This generated appropriate actuator commands, sent
wirelessly to each Animotus. Heading feedback was
provided at 1:1 mapping of user heading error to Animotus
rotation angle (saturated at ±30deg). Proximity feedback was
scaled to proximity error at approximately 1.65mm of
actuator feedback (up to 11.75mm) per meter of proximity
error. Once a user had found their current target zone, their
Animotus assumed the dormant ‘home’ pose, allowing the
user to explore the zone and listen to the audio narrative. A
large pocket on suits worn by the audience (also part of the
narrative) allowed the Animotus to be temporarily stowed, if
the user wanted to use both hands to explore a zone. The
Animotus would begin guiding the audience to their next
target zone once they left their current zone. Each audience
member was assigned a different zone order to avoid crowds
forming. All audience members were simultaneously guided
to an exit at the end of the performance, for a plot conclusion.
Flatland was experienced by 94 individuals, 15 of whom
were VI. Evaluation was achieved quantitatively (through
logged localization data) and qualitatively (via interviews).
All audience members signed a consent form approved by a
University ethics board. Numerical analysis gave insight into
a large data set from a varied pool of individuals. Though the
Animotus was initially developed under controlled laboratory
conditions, Flatland provided an opportunity for in-the-wild
testing of this technology with users who were not
necessarily focused on completing an experimental study. Of
the 94 audience members, 82% were able to locate all zones
in the space, 12% (2 VI, 9 sighted) missed one zone and 6%
(2 VI, 4 sighted) missed more than one zone.
Analysis of localization data was completed on user
trajectories between zones, referred to as paths. For each path
it was possible to calculate metrics such as average walking
speed (user distance / time elapsed between zones) and
motion efficiency (Euclidean distance / User distance
between zones). 50% efficiency indicates the user has walked
twice as far as the Euclidean distance. For each participant,
the mean of each path metric was calculated. Histograms of
this are shown in Figure 4. Both metrics show a symmetric
distribution centered on 47.5% walking path efficiency and
average walking speed of 1.125m/s. This illustrates a wide
range of participant performance. In a yet unpublished lab
study, we found a 1DOF shape changing navigation device
from a previous 2010 immersive theatre performance  to
lead to an average path efficiency of 27%, thus indicating
navigation interface improvement. Typical human walking
speed is 1.4m/s , illustrating a surprisingly small average
reduction in pace. Analysis of individual paths is underway.
Audience reaction to the Animotus varied greatly. Some
relied fully on the device, noting surprise at how intuitively
they were able to use it and commenting that without it they
would have been lost. One individual found the device too
controlling, preferring to ignore its instructions. Though no
attempt was made to make the Animotus seem like a person
or animal, audience members instilled emotional and
characterful traits, for instance referring to it as being “cute”,
“hesitant”, “a companion” and “like a pet”.
This work-in-progress paper has focused on the haptic
navigation interface used to guide audience members
between zones in the Flatland immersive theatre experience.
This project has highlighted the role haptics can play in
unifying cultural experiences across individuals of different
sensory abilities while also demonstrating the results of in-
the-wild testing of a unique shape changing device. Future
work includes in-depth analysis of the extensive data
generated from Flatland in addition to application of the
Animotus to other navigational scenarios. For example, the
device may provide haptic guidance to a distant destination
via successive waypoints or route (path) following. Attention
loading comparisons of this system with other navigation
interfaces would also be interesting.
 E. A. Abbott, Flatland: A Romance of Many Dimensions. Oxford
University Press, 1884.
 F. Hemmert, S. Hamann, and M. Löwe, “Take me by the hand: haptic
compasses in mobile devices through shape change and weight shift,”
Proceedings of the 6th Nordic Conference on Human-Computer
 J. L. Nasar and D. Troyer, “Pedestrian injuries due to mobile phone
use in public places.,” Accident; analysis and prevention, vol. 57, pp.
91–5, Aug. 2013.
 R. Velázquez, “Wearable Assistive Devices for the Blind,” Wearable
and Autonomous Biomedical Devices and Systems for Smart
Environment, pp. 331–349, 2010.
 “Apple Watch Technology,” 2015. [Online]. Available:
 S. M. Kärcher, S. Fenzlaff, D. Hartmann, S. K. Nagel, and P. König,
“Sensory augmentation for the blind.,” Frontiers in human
neuroscience, vol. 6, no. March, p. 37, Jan. 2012.
 A. A. Stanley and K. J. Kuchenbecker, “Evaluation of tactile feedback
methods for wrist rotation guidance,” IEEE Transactions on Haptics,
vol. 5, no. 3, pp. 240–251, 2012.
 S. K. Nagel, C. Carl, T. Kringe, R. Märtin, and P. König, “Beyond
sensory substitution--learning the sixth sense.,” Journal of neural
engineering, vol. 2, no. 4, pp. R13–26, Dec. 2005.
 I. Oakley and J. Park, “Did you feel something? Distracter tasks and
the recognition of vibrotactile cues,” Interacting with Computers, vol.
20, no. 3, pp. 354–363, May 2008.
 Y. Zheng and J. B. Morrell, “Haptic actuator design parameters that
influence affect and attention,” 2012 IEEE Haptics Symposium
(HAPTICS), pp. 463–470, Mar. 2012.
 J. van der Linden, Y. Rogers, M. Oshodi, A. Spiers, D. McGoran, R.
Cronin, and P. O’Dowd, “Haptic reassurance in the pitch black for an
immersive theatre experience,” in ACM Ubicomp Conference, 2011,
 R. C. Browning, E. a Baker, J. a Herron, and R. Kram, “Effects of
obesity and sex on the energetic cost and preferred speed of walking.,”
Journal of applied physiology (Bethesda, Md : 1985), vol. 100, pp.
Figure 4: Histograms of mean motion efficiency and walking
speed per participant (n = 94). Typical speed is based on .