IEEE Technology and Society Magazine - September 2020 - 72

TABLE 3. Survey questionnaire 1 (Attitude and perceptions).
Question

Top Response

Percentage

I can easily reduce my environmental impact/carbon footprint

No

64.3%

I think it is realistic to reduce my environmental impact /carbon footprint

Yes

57.1%

I know how I can minimize my impact on the environment resulting from my car use

Yes

85.7%

I usually drive when I could potentially walk

No

78.6%

Using smartphone apps in general doesn't fit into my lifestyle

No

100%

Using a parking app could help reduce the environmental impact of cars

Yes

85.7%

Generally use apps but using a parking app is too much hassle

No

71.4%

I am willing to go through the hassle of using a parking app for the greater good

Yes

85.7%

TABLE 4. Survey questionnaire 2 (Parking apps).
Top 3 Responses
Question

First

Second

Third

Do you currently use other parking apps on your phone?

No

Ringo (parking and payment
app for managed car parks )

-

What is your main motivation for using these apps?

Reduce parking time

Check for spaces

Reduce Stress

Do you use any other apps that provide a personalised
service?

Yes

No

Health apps

Do you usually check traffic/parking availability
information before starting on a journey?

Yes - Traffic

No

-

If you knew parking availability information beforehand,
would you consider changing the mode of transport?

No

Yes - if available

-

Yes

Possibly

Do you see a parking app that aims to simplify parking as No
one that encourages driving?
significantly perceivable, e.g., for someone who takes two
minutes to find parking, a reduction of half a minute may
not be noticeable. Other than this, users suggested adding
new features to improve the app, e.g., integration with a
navigation system and voice-based activation of the app.

Key Lessons from the Trial
The key benefit of ParkUs is that it provides the user with
an accurate representation of parking availability at their
destination. This requires sufficient users to populate the
map with data, users who are not attracted to use the
app if there is no data available. This was addressed in
the trial by restricting the sample to a small number of
car parks at the university campus, but the issue of how
to recruit sufficient users to make a wide-scale deployment work remains an important issue to address.
Aspects of parking behavior have been confirmed in
this trial, For example, those who arrive later in the

72

morning will spend more time cruising, as spaces are
harder to find, and further away from campus buildings.
Later arrivals would have benefitted from the app, and
could have avoided wasting time and fuel. An important
point to note here for wider deployment is that local
"peaks" in parking need may vary depending on location
and time of day. For example residential areas may see a
peak early evening as people return from work.
Small differences in cruising time were unlikely to register strongly with drivers, as they were focused on their
journey and parking. More widely of course these small
differences will accumulate to create more significant
savings across a city. In terms of the accuracy of the
cruising detection algorithm itself, it was found to show
promising generalization capability achieving an average
F1-score of 0.77 as highlighted in an earlier section. The
F1-score represents a model's classification accuracy
and is a harmonic mean of precision and recall.

IEEE TECHNOLOGY AND SOCIETY MAGAZINE

∕

SEPTEMBER 2020
IEEE Technology and Society Magazine - September 2020

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - September 2020
