IEEE Consumer Electronics Magazine - May 2018 - 95

Additionally, a coffee capsule can be present or absent in the
container. Before brewing the coffee, the operator should
insert the capsule. Once the capsule is inside the machine, the
tree is highlighted, simulating via animation the sunlight
washing over the leaves, as depicted in Figure 6. Finally, if a
critical error occurs, the machine is represented by a dead
tree, as illustrated in Figure 7, indicating that no coffee can be
brewed and that the machine requires technical service.

EXPERIMENT 1: UNDERSTANDABILITY AND AFFECT
This experiment aimed to validate if participants could read
the sensed status from the nature-inspired encoding and to
assess if BioAR is appealing. We measured errors in interpreting the BioAR tree metaphor encoding as well as engagement and overall subjective satisfaction with the machine.

(b)

(a)

FIGURE 6. The presence or absence of sun highlights encodes
whether the capsule is inserted or not. (a) A tree with no highlighting, if the capsule is absent. (b) A tree with sun highlighting,
if the capsule is present.

APPARATUS
We used the BioAR prototype with the sensor-equipped coffee machine connected to the holographic tree display. The
experimenter could set the machine to any desired state using
the available hardware interface to control the sensor data. We
made ingredients and equipment ready for participants to
brew their own coffee, including capsules with different types
of coffee, mugs and spoons, and sugar and cream. A pot of
water was available for participants to refill the machine.

PROCEDURE
We organized the study as a showcase, with a cognitive
walkthrough of the possible states of the machine. We invited participants in groups of three, with each group being
introduced to the machine and the concept as part of
the showcase. The subjects-eight males and four females,
four aged 20-29, seven 30-39, and one 40-49-received an
illustrated description of the encoding. After a brief introduction period, we asked the participants to brew themselves coffee, giving them the chance to interact with the
machine (Figure 8).
We randomly distributed the possible machine states and
their representations into 14 experimental conditions, resulting in 168 answers for each of the five variables. The participants took turns approaching the machine, which the
experimenter set to the appropriate state; we then asked them
to observe the tree hologram and mark the machine status on
a sheet of paper. Thereafter, the subjects filled out a questionnaire about their experience.

FIGURE 7. An image of a dead tree illustrates that there is a critical error in the machine.

RESULTS
The participants gave fully correct answers for all of the five
variables in each of the 14 experimental conditions in 120 out
of 168 cases. The median of these fully correct answers by participant was 11 (ranging between six and 14). Identifying the
maintenance status, determined by the number of coffees made
and shown by the presence and size of flowers, occasioned the
largest number of incorrect answers (the total score for this
variable was 130 out of 168; median = 12), followed by
categorizing the water level (total score of 155 out of 168;

FIGURE 8. Two participants in the study on understandability and
affect interact with the coffee machine and observe the consequences of their actions in the AR tree in real time.

MAy 2018

IEEE Consumer Electronics Magazine

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - May 2018