IEEE Consumer Electronics Magazine - May 2018 - 85

ON-SLOPE EXPERIMENT
For assessing the effectiveness of the
s-Helmet on the slopes, we conducted
a limited trial on actual traverse slopes
in a ski resort with six male participants (four skiers and two snowboarders), with an average age of ~30
(M = 29.833, SD = 2.124h (see Figure 3). We first properly insulated all of
the critical electronic components (i.e.,
main board) to avoid contact with the
snow. Since we were trying the s-Helmet for the first time on actual slopes,
we wanted to obtain a baseline score
for future on-slope trials with the next
versions of our prototype. As a result,
to avoid inducing fatigue in our participants that might influence our baseline
FIGURE 3. Some of our subjects testing the s-Helmet on actual ski slopes.
formation, we narrowed down our
study design to two simple conditions:
s-Helmet active and s-Helmet inactive,
while following the same procedure as before, with a
ogy used to assess approaching skiers, but also on the notiresearcher skiing behind a subject and an observer keeping a
fication system used. Our on-slope results already showed
record of how many times the participant successfully detectthat LEDs may not be ideal in many circumstances. Better
ed the position of the randomly approaching researcher.
integration into the helmet as well as increasing their intenWhen a condition was completed, participants filled out a
sity might alleviate some of these issues. Additionally, light
NASA TLX questionnaire. Once both conditions were comsensors could adaptively control the brightness of the notifipleted, we administered an SUS survey, as well as a questioncation LEDs to accommodate a wider range of lighting connaire inquiring into the subjects' skiing/snowboarding habits.
ditions (e.g., sunny versus cloudy, open sky versus forest).
We also briefly interviewed participants about their thoughts
Using audible feedback might be even more robust, yet care
and feelings regarding the prototype and the experiment.
must be taken to ensure such a system is not too noisy.
Whether audible or visible, the notification system must
gradually warn of incoming skiers, lest a warning startle
ON-SLOPE RESULTS
users and cause them to suddenly change course, which
Our on-slope results showed that with the s-Helmet active,
could actually incur-instead of prevent-a run-in with a
participants were able to accurately localize the approaching
skier from behind.
researcher 40% of the time, as opposed to 5% when the
Second, the robustness of detection would need to be caredevice was deactivated. Notably, our participants' reported
fully tuned so that one avoids too many false alarms. A comexpertise levels on the slope did not differ significantly from
bination of the two notification systems-audible and
the those of our subjects for the off-slope study. Similarly, we
visible-might offer an interesting option to improve reliabilfound no important differences between the average TLX and
ity. The LEDs could be geared toward low false negatives but
SUS scores from the on-slope trials (M TLX = 35.52%,
potentially show a few false positives, while audible notificaSD TLX = 18.04 and M SUS = 67.08, SD SUS = 17.42, respectivetions would only indicate incoming skiers with high probabilly) and the TLX and SUS scores obtained off the slope
ity (i.e., a low rate of false positives).
(M TLX = 32.7%, SD TLX = 14.65 and M SUS = 67.12, SD SUS =
13.08, respectively). The on-slope and off-slope TLX scores,
in particular, displayed almost the same workload as wearing
SkiAR: A WEARABLE AR SYSTEM FOR
a typical ski helmet (~34%). Participants' comments on the
PERSONALIZED CONTENT ON SKI RESORT MAPS
slope were somewhat similar to those for the off-slope study,
In our second prototype, we approached the question of slope
with some subjects communicating they could not see the
safety through real-time information- and experience sharing.
middle LED and one reporting he could not use his goggles
Skiing and snowboarding are undoubtedly highly social
during the trials.
activities. One important type of social activity in skiing is
decision making in a group: which piste to take next, how to
catch up with friends for lunch or après-ski, and also what
DISCUSSION
areas to avoid when going off-piste [10].
The goal of the s-Helmet is to gradually warn users of skiConventionally, skiers and snowboarders consult ski resort
ers approaching from behind. Clearly, the usefulness of
paper maps or large-scale maps mounted along the slopes,
such an approach critically depends not only on the technolMAy 2018

IEEE Consumer Electronics Magazine

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