Inter
national
J
our
nal
of
Electrical
and
Computer
Engineering
(IJECE)
V
ol.
7,
No.
6,
December
2017,
pp.
3602
–
3612
ISSN:
2088-8708
3602
I
ns
t
it
u
t
e
o
f
A
d
v
a
nce
d
Eng
ine
e
r
i
ng
a
nd
S
cie
nce
w
w
w
.
i
a
e
s
j
o
u
r
n
a
l
.
c
o
m
SmartBik
e:
an
IoT
Cr
o
wd
Sensing
Platf
orm
f
or
Monitoring
City
Air
P
ollution
Fulvio
Cor
no
1
,
T
eodor
o
Montanar
o
2
,
Carmelo
Miglior
e
3
,
and
Pino
Castr
ogio
v
anni
4
1,2,3
Department
of
Control
and
Computer
Engineering,
Politecnico
di
T
orino,
Italy
4
SW
ARM
Joint
Open
Lab,
TIM,
T
orino,
Italy
Article
Inf
o
Article
history:
Recei
v
ed:
Apr
5,
2017
Re
vised:
Sep
7,
2017
Accepted:
Sep
20,
2017
K
eyw
ord:
Internet
of
Things
Smart
City
Cro
wd
Sensing
bik
e
pollution
location
ABSTRA
CT
In
recent
years,
the
Smart
Cit
y
concept
is
emer
ging
as
a
w
ay
to
increase
ef
ficienc
y
,
reduce
costs,
and
impro
v
e
the
o
v
erall
quality
of
cit
izen
life.
The
rise
of
Smart
City
so-
lutions
is
e
ncouraged
by
the
increasing
a
v
ailability
of
Internet
of
Things
(IoT)
de
vices
and
cro
wd
sensing
technologies.
This
pape
r
presents
an
IoT
Cro
wd
Sensing
platform
that
of
fers
a
set
of
services
to
citizens
by
e
xploiting
a
netw
ork
of
bic
ycles
as
IoT
probes.
Based
on
a
surv
e
y
conducted
to
identify
the
most
interesting
bik
e-enabled
ser
-
vices,
the
SmartBik
e
platform
pro
vides:
real
time
remote
geo-location
of
users’
bik
es,
anti-theft
service,
information
about
tra
v
eled
route,
and
air
pollution
monitoring.
The
proposed
SmartBik
e
platform
is
composed
of
three
m
ain
components:
the
SmartBik
e
mobile
sensors
for
data
collection
installed
on
the
bic
ycle;
the
end-user
de
vices
imple-
menting
the
user
interf
ace
for
geo-location
and
anti-the
ft;
and
the
SmartBik
e
central
serv
ers
for
storing
and
processing
detected
data
and
pro
viding
a
web
interf
ace
for
data
visualization.
The
suitability
of
the
platform
w
as
e
v
aluated
through
the
implementa-
tion
of
an
initial
prototype.
Results
demonstrate
that
the
proposed
SmartBik
e
platform
is
able
to
pro
vide
the
stated
services,
and,
in
addition,
that
the
accurac
y
of
the
acquired
air
quality
measurements
is
compatible
with
the
one
pro
vided
by
t
he
of
ficial
en
viron-
mental
monitoring
system
of
the
city
of
T
urin.
The
described
platform
will
be
adopted
within
a
project
promoted
by
t
he
city
of
T
urin,
that
aims
at
helping
people
making
their
mobility
beha
vior
more
sustainable.
Copyright
c
2017
Institute
of
Advanced
Engineering
and
Science
.
All
rights
r
eserved.
Corresponding
A
uthor:
T
eodoro
Montanaro
Politecnico
di
T
orino
Corso
Duca
de
gli
Abruzzi,
24
-
T
orino,
Italy
10129
+390110907191
teodoro.montanaro@polito.it
1.
INTR
ODUCTION
Recently
,
the
concept
of
Smart
City
has
emer
ged
as
a
w
ay
to
“e
xploit
the
Information
and
Communi-
cation
T
echnologies
(ICT)
in
making
better
use
of
the
public
resources,
increase
the
quality
of
services
of
fered
to
the
citi
zens
and,
in
turn,
the
quality
of
life
in
urban
areas,
while
reducing
the
operational
costs
of
the
public
administrations”
[1].
In
Smart
Cities,
a
broad
netw
ork
of
IoT
sensors
(such
as
smartphones,
smart
v
ehicles,
or
surv
eillance
cameras)
is
spread
across
the
city
with
the
aim
of
collecting
data
to
foster
the
de
v
elopment
of
inno
v
ati
v
e
applications
for
citizens,
companies,
and
public
administrations
[2].
One
of
the
w
ays
used
to
en-
hance
the
contrib
ution
of
IoT
sensors
to
Smart
Cities
is
by
e
xploiting
cro
wd
sensing
solutions.
Cro
wd
sensing
is
an
alt
ernati
v
e
to
the
traditional
w
ay
of
g
athering
and
del
i
v
ering
information
from
and
to
the
en
vironment:
sensing
is
“distrib
uted
across
a
lar
ge
number
of
(often
mobile)
indi
viduals
rather
than
by
an
y
indi
vidual
or
set
of
sensors
in
fix
ed
locations”
[3].
Essentially
,
e
v
ery
de
vice
o
wned
by
a
person
(e.g.,
smartphone,
smartw
atch,
etc.)
can
be
used
to
acquire
geo-located
data
thus
permitting
spatial
and
temporal
resoluti
o
n
impro
v
ements
in
services
pro
vided
by
a
smart
city
.
J
ournal
Homepage:
http://iaesjournal.com/online/inde
x.php/IJECE
I
ns
t
it
u
t
e
o
f
A
d
v
a
nce
d
Eng
ine
e
r
i
ng
a
nd
S
cie
nce
w
w
w
.
i
a
e
s
j
o
u
r
n
a
l
.
c
o
m
,
DOI:
10.11591/ijece.v7i6.pp3602-3612
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISSN:
2088-8708
3603
The
goal
of
this
paper
is
the
design
and
the
de
v
elopment
of
an
IoT
cro
wd
sensing
platform
(the
SmartBik
e
platform)
able
to
monitor
air
city
conditions
while
pro
viding
services
that
are
interesting
for
both
users
and
institutions.
This
platform
will
be
adopted
within
the
OpenAgor
`
a
project
[4],
one
of
the
proposals
selected
by
the
city
of
T
urin
for
the
T
orino
Li
ving
Lab
Campidoglio
e
xperimentation
[5].
In
v
olving
dif
ferent
partners
(T
urin
TIM
Joint
Open
Lab,
Politecnico
di
T
orino,
and
tw
o
startups,
Mo
v
e
P
lus
[6]
and
Pon
yzero
[7]),
the
Open
Agor
`
a
project
aims
at
de
v
eloping
and
testing
solutions
for
helping
people
to
mak
e
their
mobility
beha
vior
more
sustainable.
In
addition,
it
aims,
at
the
same
time,
at
pro
viding
data
and
tools
that
can
be
used
by
the
city
institutions
to
enhance
the
o
v
erall
quality
of
life
of
their
citizens.
Due
to
the
e
xtensi
v
e
use
of
bic
ycles
observ
ed
in
literature
for
monitoring
smart
cities
([8,
9,
10,
11,
12]),
the
bic
ycle
w
as
selected
as
a
cro
wd
sensing
probe
for
monitoring
the
city
en
vironment.
The
de
v
elopment
of
the
proposed
platform
w
as
di
vided
into
four
main
steps.
At
first,
a
surv
e
y
w
as
conducted
to
identify
the
most
interesting
bik
e-enabled
features
for
users.
This
step
is
based
on
the
observ
ation
reported
by
Alam
et
al.
[13]
about
the
usefulness
of
in
v
olving
users
in
designing
ne
w
services:
authors
demonstrate
that
user
in
v
olv
ement
in
designing
ne
w
services
f
acilitates
the
de
v
elopment
of
better
and
dif
ferentiated
ne
w
services
that
match
e
xactly
customer
needs.
Then,
in
the
second
phase,
the
four
most
preferred
features
emer
ging
from
the
surv
e
y
results
were
selected
to
inform
the
definition
of
the
services
pro
vided
by
t
he
platform.
The
selected
features
are
the
follo
wing:
a)
real
time
remote
geo-location
of
user
bik
es,
b)
anti-theft
service,
c)
information
about
tra
v
eled
route
(distance,
duration,
and
rise),
and
d)
air
pollution
monitoring.
Based
on
them,
in
this
phase,
the
architec-
ture
of
an
IoT
cro
wd
sensing
platform
w
as
designed.
After
that,
an
initial
prototype
of
the
presented
platform
w
as
implemented
in
the
third
phase
and
used
in
the
forth
phase
to
test
the
feasibility
of
the
approach
and
the
suitability
of
the
platform.
It
w
as
demonstrated
that
the
proposed
SmartBik
e
platform
is
able
to
pro
vide
the
designed
servic
es,
and,
in
addition,
that
the
air
pollution
measures
pro
vided
by
the
platform
are
compatible
with
the
ones
pro
vided
by
the
of
ficial
en
vironmental
monitoring
system
of
the
city
of
T
urin.
The
remainder
of
the
paper
is
or
g
anized
as
follo
ws:
Section
2.
analyzes
related
w
orks,
while
Sec
tion
3.
presents
the
results
of
the
user
surv
e
y
.
Section
4.
describes
the
proposed
architecture
and
Section
5.
analyzes
the
first
de
v
eloped
prototype.
Finally
,
results
obtained
by
testing
the
feasibility
of
the
platform
are
discussed
in
Section
6.
and
Section
7.
concludes
the
paper
and
discusses
future
w
orks.
2.
RELA
TED
W
ORKS
The
current
section
presents
e
xisting
en
vironmental
monitoring
systems
and,
specifically
,
air
qual
ity
detection
solutions
based
on
IoT
technologies.
P
articular
attention
is
paid
to
the
services
for
citizens
pro
vided
by
each
presented
solution.
En
vironmental
monitoring
and
air
quality
detection
systems
ha
v
e
been
subject
of
e
xtensi
v
e
st
ud
i
es
in
the
literature
([14,
15,
16]).
Mehta
et
al.
[17],
for
e
xample,
de
v
eloped
a
cloud
based
air
quality
detection
system
that
analyzes
data
for
pro
viding
atmospheric
quality
measurements
to
the
user
in
real
time.
The
y
proposed
the
architecture
of
a
AMS
(Air
Monitoring
Sensor)
system
that
consists
of
a
de
vice
equipped
with
nine
sensors
placed
in
strate
gic
locations
across
the
city
.
The
de
vice
is
responsible
for
collecting
data
about
the
toxicity
of
the
air
,
informing
users
about
the
air
quality
of
the
area
in
which
the
y
are
mo
ving
and
suggesting
alternati
v
e
routes.
Xiaojun
et
al.
[18]
presented
an
IoT
-based
air
pollution
monitoring
and
forecasting
system.
The
designed
system
is
based
on
lo
w
cost
air
quality
monitoring
stations
that
can
be
laid
out
in
sensible
points
of
the
city
.
These
stations
enhance
the
netw
ork
of
sensors
spread
all
o
v
er
the
city
and
increase
the
precision
of
re
v
ealed
air
quality
information.
The
prese
n
t
ed
solution
does
not
actually
supply
an
y
service
to
users,
b
ut
it
constitutes
a
real
and
ef
fecti
v
e
decision-making
basis
for
emer
genc
y
response
that
can
be
used
to
de
v
elop
and
pro
vide
ne
w
future
services.
In
addition,
dif
ferent
other
w
orks
were
found
in
literature
aiming
at
monitoring
ambient
condi
tions
using
cro
wd-sensing
based
de
vices,
such
as
bic
ycles,
and
complementing
local
en
vironmental
monitoring
sta-
tions.
Liu
et
al.
[8,
10]
presented
a
platform
that
e
xploits
a
custom-made
sensor
box
mounted
on
public
shared
bic
ycles
to
monitor
ai
r
pollution.
The
y
demonstrate
that
these
de
vices
can
be
used
to
impro
v
e
the
accurac
y
of
the
e
xisting
local
en
vironmenta
l
monitoring
stations.
The
air
quality
of
the
entire
city
is
presented,
in
both
w
orks,
on
a
website
and
is
a
v
ailable
to
e
v
ery
citizens
and/or
institutions.
Furthermore,
Zeiger
et
al.
[9]
described
a
system
for
en
vironmental
monitoring
based
on
partici
patory
sensing.
Mobile
en
vironmental
sensors
carried
by
citizens
(pedestrians
or
c
yclists)
are
used
to
measure
pollu-
tant
concentrations.
Acquired
data
are
then
used
to
b
uild
a
pollution
map.
Another
interesting
system,
similar
SmartBik
e:
an
IoT
Cr
owd
Sensing
Platform
for
Monitoring
City
Air
P
ollution
(T
eodor
o
Montanar
o)
Evaluation Warning : The document was created with Spire.PDF for Python.
3604
ISSN:
2088-8708
to
the
one
presented
in
this
paper
,
is
the
SensorW
ebBik
e,
proposed
by
V
agnoli
et
al.
[11].
The
main
component
of
their
system
is
a
de
vice
equipped
with
air
quality
sensors
(such
as,
noise,
humidity
,
temperature,
C
O
,
C
O
2
,
O
3
,
N
O
2
,
and
C
H
4
sensors)
that
monitors
the
en
vironment
by
using
users’
bic
ycles
as
probes.
Ho
we
v
er
,
the
main
dif
ference
with
the
platform
presented
in
this
paper
is
related
to
the
services
directly
supplied
to
citizens.
The
SensorW
ebBik
e
pro
vides
only
a
web
application
that
sho
ws
a
pollution
map
of
the
city
to
citizens
without
other
user
-related
services
such
as
anti-theft.
Finally
,
V
elasco
et
al.
[19]
present
a
mobile
wireless
sensor
net-
w
ork
system
aiming
at
complementing
the
already
e
xisting
of
ficial
air
quality
monitoring
systems
of
the
city
of
T
orino.
Some
tests
were
carried
out
by
mounting
the
proposed
system
on
bic
ycles
in
order
to
increase
their
mobility
.
The
main
dif
fere
n
c
e
from
the
one
described
in
the
current
paper
is
that
their
system
does
not
of
fer
direct
services
to
citizens
and,
specifically
,
to
c
yclists.
3.
USER
SUR
VEY
As
the
first
step
in
de
v
eloping
the
SmartBik
e
platform,
an
online
surv
e
y
w
as
conducted
with
the
aim
of
identifying
the
most
interesting
bik
e-enabled
features.
288
persons
were
in
v
olv
ed
among
Politecnico
di
T
orino
students
and
TIM
emplo
yees
that
usually
mo
v
e
around
the
city
of
T
urin
(Italy)
riding
their
bik
es.
The
surv
e
y
has
been
accessible
for
tw
o
months
and
consist
ed
of
10
questions
di
vided
into
three
main
sections.
The
first
section
aimed
at
acquiring
user
demographic
and
habits
informat
ion.
The
second
section,
instead,
had
the
objecti
v
e
of
analyzing
user
preferences
for
the
future
design
of
aesthetic
characteristics
of
the
proposed
solution.
The
questions
ask
ed
users
to
e
xpress
their
preferences
about
the
preferred
type
of
bik
e
(e.g.,
mountain
bik
e,
city
bik
e,
etc.)
and
the
most
desired
bik
e
accessories
(e.g.,
bask
et,
bagg
age
holder
,
etc.).
Finally
,
the
third
section
w
as
related
to
the
identification
of
the
most
interesting
services
for
c
yclists
in
the
conte
xt
of
a
smart
city
.
The
language
used
within
the
surv
e
y
w
as
Italian,
results
were
then
translated
for
the
purposes
of
this
paper
.
3.1.
Demographic
inf
ormation
about
inter
viewees
Aiming
at
recruiting
users
that
usually
mo
v
e
around
the
city
of
T
urin
(Italy)
riding
their
bik
es,
500
promotional
tags
with
a
printed
QR
code
(which
pointed
to
the
surv
e
ys
link)
were
attached
to
the
bik
es
park
ed
in
the
Politecnico
di
T
orino
courtyard.
Moreo
v
er
,
an
email
promoting
the
surv
e
y
w
as
sent
to
TIM
emplo
yees.
A
population
of
288
people
replied
to
the
study
,
with
221
males
and
67
females.
Most
participants
were
aged
in
the
interv
al
“36
-
50”
(Figure
1).
Furthermore,
Figure
2
s
ho
ws
the
frequenc
y
with
which
users
declared
to
use
their
bic
ycles:
most
of
the
intervie
wees
(81%)
commonly
use
their
bik
es
se
v
eral
times
a
month,
and
50%
at
least
weekly
.
Figure
1.
User
Surv
e
y:
age
distrib
ution
Figure
2.
User
Surv
e
y:
frequenc
y
of
bic
ycle
usage
3.2.
Pr
eferr
ed
type
of
bik
e
and
most
used
bik
e
de
vices
In
the
second
section
of
the
surv
e
y
,
respondents
were
ask
ed
to
select
the
preferred
type
of
bik
e
among
the
follo
wing
four
types:
IJECE
V
ol.
7,
No.
6,
December
2017:
3602
–
3612
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISSN:
2088-8708
3605
b
udget
bik
e
city
bik
e
mountain
bik
e
racing
bik
e.
Results
re
v
eals
that
the
most
preferred
type
of
bik
e
is
the
city
bik
e
(45.8%),
follo
wed
by
the
mountain
bik
e
(26.7%),
the
racing
bik
e
(22.0%)
and,
finally
the
b
udget
bik
e
(5.5%).
Furthermore,
users
were
ask
ed
to
declare
if
the
y
usually
desire
one
or
more
of
the
follo
wing
bik
e
accessories:
bagg
age
holder
bask
et
frontal
light
spok
e
light.
In
this
case,
results
re
v
eals
that
the
most
preferred
de
vice
is
the
frontal
light
(87.7%),
follo
wed
by
the
spok
e
light
(54.8%),
the
bagg
age
holder
(46.6%)
and,
finally
,
the
bask
et
(28.8%).
3.3.
Most
inter
esting
bik
e-enabled
featur
es
In
the
final
section
of
the
surv
e
y
,
users
were
ask
ed
to
select
up
to
six
possible
future
high
tech
bik
e
impro
v
ements
out
of
those
listed
in
T
able
1.
The
reported
v
alues
represent
the
percentage
of
users
that
selected
each
feature.
As
can
be
observ
ed
from
table
1,
the
six
most
requested
features
are:
real
time
geo-location
detec
tion
of
the
bik
e
in
case
of
loss
or
theft,
anti-theft
service
which
can
send
notifications
to
an
end-user
de
vice,
infor
-
mation
about
tra
v
eled
route
(tra
v
eled
distance,
duration,
dif
ference
in
altitude),
air
pollution
le
v
el
of
tra
v
eled
roads,
GPS
na
vig
ation
de
vice,
and
information
about
the
speed.
T
able
1.
List
of
possible
technological
bik
e
impro
v
ements
with
percentage
of
users
that
selected
each
of
them
F
eatur
e
P
er
centage
of
selections
(o
v
er
288
users)
Real
time
remote
geo-location
detection
of
the
bik
e
in
case
of
loss
or
theft
67.72%
Anti-theft
feature
which
can
send
notification
to
an
end-user
de
vice
64.67%
Information
about
tra
v
eled
routes
(tra
v
eled
distance,
duration,
dif
ference
in
altitude)
47.48%
Air
pollution
le
v
el
of
tra
v
eled
roads
42.07%
GPS
na
vig
ation
de
vice
31.49%
A
v
erage,
minimum,
and
maximum
speed
30.25%
Automatic
call
for
assistance
in
case
of
accidents
(e-call)
29.70%
Bic
ycle
maintenance
status
23.79%
Information
about
high
injury
risk
roads
22.28%
Burned
calories
13.09%
T
raf
fic
information
13.05%
Audio
instructions
about
na
vig
ation
11.81%
Heartbeat
monitoring
9.69%
Reminders
based
on
location
on
the
route
7.98%
Point
of
interest
7.42%
Information
about
not
accessible
roads
6.93%
SmartBik
e:
an
IoT
Cr
owd
Sensing
Platform
for
Monitoring
City
Air
P
ollution
(T
eodor
o
Montanar
o)
Evaluation Warning : The document was created with Spire.PDF for Python.
3606
ISSN:
2088-8708
4.
ARCHITECTURE
In
the
second
phase
of
the
SmartBik
e
platform
de
v
elopment,
the
four
most
requested
feat
ures
resulting
from
the
surv
e
y
were
selected.
Based
on
them,
tw
o
dif
ferent
services
were,
then,
designed
and
the
architecture
of
the
platform
w
as
de
vised.
The
selected
features
are:
(a)
real
time
remote
geo-location
detection
of
the
users’
bik
es;
(b)
anti-theft;
(c)
information
about
tra
v
eled
route
(distance,
duration,
and
rise);
(d)
air
pollution
monitoring.
4.1.
Scenario
Starting
from
the
surv
e
y
results,
the
follo
wing
enabling
scenario
w
as
designed
and
used
to
inform
the
definition
of
the
SmartBik
e
platform
pro
vided
services.
The
scenario
concerns
the
use
of
the
system
by
a
typical
c
yclist
(named
Ann)
whose
bik
e
is
stolen:
Ann
is
at
her
home
and
wants
to
use
her
bik
e
to
go
to
work
by
avoiding
ar
eas
with
too
muc
h
pollution.
Consequently
,
she
tak
es
her
smartphone
,
opens
the
SmartBik
e
app
and,
by
using
the
b
uilt
in
maps
c
hooses
the
best
r
oute
to
go
to
her
works.
Then,
she
enter
s
her
gar
a
g
e
to
tak
e
the
bik
e
,
she
deactivates
the
anti-theft
service
and
then
she
goes
working
by
bik
e
.
When
Ann
arrives
at
her
workplace
,
she
parks
the
bik
e
in
nearby
park,
activates
the
anti-theft
system
by
using
her
smartphone
and
enter
s
her
of
fice
.
While
she
is
working
,
someone
tries
to
steal
her
bik
e
,
so
a
notification
is
sent
to
her
smartphone
.
Immediately
she
e
xits
her
of
fice
to
scar
e
the
thief
,
b
ut
he
is
alr
eady
gone
.
Howe
ver
,
by
using
the
SmartBik
e
app
she
can
locate
her
bik
e
in
r
eal
time
,
so
she
contacts
the
police
and
the
y
finally
r
eco
ver
her
bik
e
.
4.2.
Pr
o
vided
ser
vices
Looking
at
the
scenario,
the
follo
wing
tw
o
main
services
were
identified
as
the
ones
that
should
be
pro
vided
by
the
SmartBik
e
platform:
1.
the
“city
monitoring”
service
g
athers
the
features
(a,
c
and
d
of
the
first
list
of
this
section)
aimed
at
collecting
and
sho
wing
air
pollution
information
of
areas
tra
v
eled
by
in
v
olv
ed
c
yclists.
Considering
that
air
pollution
can
be
influenced
by
temperature,
relati
v
e
humidity
and
barometric
pressure,
those
data
should
be
acquired
through
appropriate
sensors
in
addition
to
the
air
pollution
information.
The
collected
data
should
be
geo-located
and
periodically
sent
to
SmartBik
e
central
serv
ers
able
to
store
it.
Lik
e
wise,
the
platform
should
pro
vide
a
web
map
sho
wing
the
tra
v
eled
routes
and
the
le
v
el
of
pollution
of
the
areas
of
the
city
monitored
by
a
v
ailable
bik
es.
These
services
should
be
accessible
into
tw
o
dif
ferent
modes:
a
“personal”
mode
sho
wing
only
user
related
information
to
logged
users
(e.g.,
the
position
of
the
o
wned
bik
es)
and
a
“public”
mode
sho
wing
aggre
g
ated
information
obtained
by
mer
ging
data
collected
by
each
user;
2.
the
“anti-theft”
service,
instead,
g
athers
the
anti-theft
feature
(b
of
the
list
reported
at
the
be
ginning
of
this
section).
The
SmartBik
e
platform
should
allo
w
an
authenticated
user
to
enable/disable
an
anti-theft
service
that,
by
monitori
ng
mo
v
ements
of
the
bik
e,
should
generate
a
notification
whene
v
er
a
thief
tries
to
steal
the
bi
k
e.
In
addition,
the
platform
should
pro
vide
real
time
information
about
the
bik
e
location
and
the
de
vice
status
(e.g.,
battery
le
v
el).
Figure
3
sho
ws
the
designed
logical
architecture
of
the
SmartBik
e
platform.
It
is
composed
of
three
main
components:
the
SmartBik
e
de
vices,
the
end-user
de
vices
(e.g.,
smartphones
and
t
ablets),
and
the
Smart-
Bik
e
central
serv
ers.
4.3.
Ar
chitectur
e
design
The
SmartBik
e
de
vices
block
represents
the
IoT
objec
ts
that
are
mounted
on
the
bic
ycles.
The
y
should
be
autonomous
(i.e.,
al
w
ays
acti
v
e
and
connected
to
the
Internet
e
v
en
when
the
user
smartphone/tablet
is
not
close
to
them)
and
should
pro
vide
the
follo
wing
functionalities
to
support
t
he
tw
o
services
described
in
Section
4.2.:
IJECE
V
ol.
7,
No.
6,
December
2017:
3602
–
3612
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISSN:
2088-8708
3607
Figure
3.
High
le
v
el
architecture
design
of
the
SmartBik
e
platform
en
vironmental
monitoring
f
un
c
tionality
able
to
collect
data
about
air
pollution,
temperature,
relati
v
e
humidity
and
barometric
pressure;
bik
e
status
monitoring
functionality
able
to
detect
an
y
mo
v
ement
of
the
bik
e
while
the
anti-theft
service
is
enabled;
data
synchronization
functionality
able
to
periodically
send
acquired
data
to
the
SmartBik
e
central
serv
ers.
When
the
anti-theft
service
is
enabled
the
information
should
be
sent
in
real
time
to
the
SmartBik
e
central
serv
ers;
location
functionality
used
to
geo-tag
acquired
information
and
locate
the
bik
e
in
case
of
theft
or
loss;
communication
functionality
used
to
pro
vide
interactions
with
end-user
de
vices.
The
end-user
de
vices
block
represents
the
smartphones
or
tablets
on
which
a
dedicated
SmartBik
e
application
is
installed.
This
app
permits
the
follo
wing
functionalities
with
the
SmartBik
e
de
vices
and
the
SmartBik
e
central
serv
ers:
authenticate
the
user
and
enable
interactions
with
her
o
wn
SmartBik
e
de
vices;
enable/disable
anti-theft
service;
generate
theft
alert
notifications
whene
v
er
it
is
informed
about
a
theft
by
the
SmartBik
e
central
serv
ers;
present
the
current
status
information
(i.e.,
g
as
concentration,
temperature,
relati
v
e
humidity
,
barometric
pressure
and
battery
le
v
el)
about
all
SmartBik
e
de
vices
o
wned
by
the
user;
visualize
the
location
of
the
de
vice
on
a
map;
sho
w
information
about
tra
v
eled
route.
Finally
,
the
SmartBik
e
central
ser
v
ers
represent
one
or
more
back-end
serv
ers
that
pro
vide
three
dif
ferent
kinds
of
functionalities
as
parts
of
both
the
tw
o
described
services:
1.
a
data
collection
functionality
,
able
to
periodically
recei
v
e
data
sent
by
SmartBik
e
de
vices
o
wned
by
dif
ferent
users.
These
data
are
useful
for
both
final
users
and
city
institutions;
2.
a
web
interf
ace
similar
to
the
one
pro
vided
by
the
app
run
on
end-user
de
vices
to
supply
a)
a
map
to
geo-locate
in
real
time
the
user
bik
es,
b)
a
map
based
on
historical
information
to
sho
w
the
air
pollution
conditions
of
the
areas
tra
v
eled
by
in
v
olv
ed
c
yclists,
c)
information
about
personal
tra
v
eled
route.
This
interf
ace,
and
specifically
,
the
functionalities
described
in
b)
and
c)
will
be
interesting
for
both
c
yclists
and
city
institutions;
3.
an
anti-theft
functionality
that
a)
redirects
theft
notifications
sent
by
SmartBik
e
de
vi
ces
to
the
right
end-
user
de
vices
and
b)
sends
an
anti-theft
notification
if
no
data
are
recei
v
ed
from
a
SmartBik
e
de
vice
for
a
certain
amount
of
time.
SmartBik
e:
an
IoT
Cr
owd
Sensing
Platform
for
Monitoring
City
Air
P
ollution
(T
eodor
o
Montanar
o)
Evaluation Warning : The document was created with Spire.PDF for Python.
3608
ISSN:
2088-8708
5.
PR
O
T
O
TYPE
In
order
to
e
v
aluate
the
suitability
of
the
platform,
an
initial
prototype
of
the
presented
archite
cture
has
been
implemented.
As
an
initia
l
prototype,
the
main
assumption
considered
in
the
follo
wing
description
is
that
a
user
o
wns
only
one
bik
e
and
one
end-user
de
vice
(e.g.,
smartphone
and
tablet).
Ho
we
v
er
,
the
prototype
can
be
easily
e
xtended
to
ca
ses
in
which
a
single
user
o
wns
more
than
one
bic
ycle
and/or
more
end-user
de
vices.
In
addition,
in
this
prototype,
the
air
pollution
w
as
monitored
through
the
carbon
monoxide
concentration,
only
.
The
follo
wing
subsections
illustrate
the
details
of
the
components
that
compose
the
presented
plat-
form.
5.1.
SmartBik
e
de
vices
Figure
4
sho
ws
the
implemented
SmartBik
e
de
vice
prototype.
It
w
as
implemented
using
the
STM32
Nucleo
L476RG
board
[20]
equipped
with
the
ARM
mbed
3.0
Operating
System
[21]
as
control
board.
The
board
contains
the
basic
components
(i.e.,
the
CPU,
memory
and
some
ports)
b
ut
it
can
be
easily
e
xtended
with
a
lar
ge
number
of
specialized
application
hardw
are
add-ons
and
shields.
In
this
w
ork
it
w
as
e
xpanded
by
adding
the
follo
wing
shields:
Adafruit
FON
A808
[22],
as
GSM/GPRS
and
GPS
module
for
transmitting
data
to
the
SmartBik
e
central
serv
ers
and
pro
vide
the
location
of
the
bik
es;
ST
X-Nucleo-IDB05A1
[23],
as
Bluetoot
h
Lo
w
Ener
gy
4.1
shield
for
pro
viding
connections
with
near
end-user
de
vices;
ST
X-Nucleo-IKS01A1
[24],
as
en
vironmental
and
motion
sensor
s
hield
for
acquiring
temperature,
hu-
midity
,
pressure
and
motion
(acquired
through
the
accelerometer);
Nemoto
N
AP-505
[25]
with
T
e
xas
Instruments
LMP91002
Analog
Front
End
(AFE)
[26],
as
electro-
chemical
CO
sensor
chosen
for
its
lo
w
po
wer
consumption
and
good
reading
accurac
y
at
a
reasonable
cost.
The
de
vice
is
autonomous:
a
battery
guarantees
its
po
wer
supply
and
the
GSM/GPRS
module
guarantees
its
connecti
vity
functions
e
v
en
when
it
is
not
connected
to
a
smartphone/tablet.
As
a
requirement
of
this
prototype,
a
10
Ah
LiPo
battery
w
as
chosen
to
let
the
de
vice
be
acti
v
e
for
48
hours
while
collecting
data
with
an
int
erv
al
of
12
seconds.
5.2.
End-user
de
vices
Figure
5
sho
ws
tw
o
screenshots
of
the
designed
end-user
prototype
application.
The
application
w
as
de
v
eloped
for
Android
de
vices
running
at
least
Android
5.0
and
w
as
de
v
eloped
using
the
Android
Studio
IDE.
The
implemented
user
interf
ace
is
minimal
and
is
composed
of
tw
o
acti
vities
pro
viding
dif
ferent
functionalities.
The
main
acti
vity
is
sho
wn
in
Figure
5a.
It
implements
the
follo
wing
functionalities:
it
pro
vides
a
b
utton
to
connect
the
end-user
de
vice
to
the
hardw
are
de
vice
via
a
Bluetooth
Lo
w
Ener
gy
connection;
if
connected
to
the
hardw
are
de
vice,
it
pro
vides
the
current
status
of
the
hardw
are
de
vice
sho
wing
the
re
v
ealed
carbon
monoxide,
temperature,
relati
v
e
humidity
and
pressure;
if
connected
to
the
hardw
are
de
vice,
it
pro
vides
a
b
utton
to
acti
v
ate
or
deacti
v
ate
the
anti-theft
service.
The
second
acti
vity
,
i.e.,
the
map
acti
vity
,
is
sho
wn
in
Figure
5b
and
pro
vi
d
e
s
a
map
sho
wing
the
current
location
of
the
bik
e.
In
addition,
tw
o
background
services
were
impl
emented:
the
first
one
pro
vides
an
interf
ace
for
re-
cei
ving
notifications
from
the
SmartBik
e
central
serv
ers
in
case
of
t
heft,
while
the
s
econd
one
manages
the
connections
needed
to
periodically
send
collected
data
to
the
SmartBik
e
central
serv
ers.
IJECE
V
ol.
7,
No.
6,
December
2017:
3602
–
3612
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISSN:
2088-8708
3609
Figure
4.
The
SmartBik
e
de
vice
prototype
Figure
5.
Screenshots
of
the
designed
proto-
typal
application
5.3.
SmartBik
e
central
ser
v
ers
T
o
pro
vide
the
designed
services,
a
solution
inte
grating
a
back-end
and
a
front-end
platform
w
as
adopted
in
impl
ementing
the
SmartBik
e
central
serv
ers
block.
SiteWhere
[27],
an
open
source
IoT
serv
er
platform,
w
as
used
as
back-end
to
collect
and
store
all
the
data
sent
by
both
Smart
Bik
e
de
vices
and
end-user
de
vices
and
to
pro
vide
this
data
to
the
mobile
and
the
web
applications.
It
collected:
a)
air
pollution
information,
b)
location
information,
c)
bik
e
motion
(acquired
through
accelerometer),
d)
SmartBik
e
de
vices’
statuses.
The
web
application,
acting
as
front-end
platform,
w
as
implemented
based
on
Meteor
[28],
a
full-
stack
Ja
v
aScript
platform
for
de
v
eloping
modern
web
and
mobile
applications.
The
implemented
Meteor
application
grants
a)
access
to
a
map
sho
wing
tra
v
eled
routes
and
the
le
v
el
of
pollution
of
the
areas
of
the
city
monitored
by
a
v
ailable
bik
es,
and
b)
the
generation
of
all
the
notifications
sent
to
users
to
report
theft
attempts.
The
solution
w
as
deplo
yed
on
a
free
Amazon
EC2
[29]
instance.
Figure
6
sho
ws
the
map
reported
in
the
web
interf
ace
to
let
the
user
kn
o
w
her
bik
e
current
l
o
c
ation
and
the
recently
tra
v
eled
route.
Figure
6.
The
position
of
the
bik
e
can
be
remotely
visualized
on
a
map
with
the
tra
v
eled
route
6.
RESUL
TS
T
o
e
v
aluate
t
h
e
feasibility
of
the
approach,
the
suitability
of
the
platform
and
the
accurac
y
of
the
implemented
prototype,
the
whole
prototype
has
been
subjected
to
a
test
phase.
A
v
olunteer
c
yclist
w
as
in
vited
to
reproduce
the
sample
scenario
reported
in
section
4.:
the
implemented
prototype
of
the
SmartBik
e
de
vice
w
as
placed
inside
the
bask
et
of
his
bik
e
and
the
Android
application
w
as
i
nstalled
on
his
smartphone
running
SmartBik
e:
an
IoT
Cr
owd
Sensing
Platform
for
Monitoring
City
Air
P
ollution
(T
eodor
o
Montanar
o)
Evaluation Warning : The document was created with Spire.PDF for Python.
3610
ISSN:
2088-8708
Figure
7.
Map
of
the
area
of
T
urin
(Italy)
monitored
within
the
test
Android
6.0.1
(a
Karbonn
Sparkle
V
smartphone).
The
e
xperiment
w
as
conducted
on
the
26th
of
May
2016
for
30
minutes
between
14:00
and
14:30.
The
user
went
through
the
area
of
T
urin
(Italy)
sho
wn
in
Figure
7
and
located
near
Politecnico
di
T
orino.
Air
pollution
information
were
collected
simultaneously
usi
n
g
a)
the
implemented
platform
prototype
and
b)
the
closest
ARP
A
(Ita
lian
acron
ym
of
“Re
gional
En
vironmental
Protection
Agenc
y”)
city
monitoring
station
situated
about
3
km
f
ar
from
the
area
of
e
xperiment
(it
is
precisely
located
in
V
ia
della
Consolata
in
T
urin,
Italy).
Moreo
v
er
,
the
anti-theft
service
w
as
tested
by
trying
to
mo
v
e
the
park
ed
bik
e
to
simulate
a
theft
attempt
without
the
presence
of
the
bik
e
o
wner
.
The
analysis
of
collected
data
actually
demonstrated
that
the
implemented
prototype
w
as
able
to
pro-
vide
all
the
services
described
in
the
pre
vious
sections:
a
notification
arri
v
ed
on
the
user
smartphone
after
a
fe
w
seconds
from
the
be
ginning
of
the
theft
attempt;
the
CO
v
alue
measured
by
the
SmartBik
e
de
vice
w
as
compatible
wi
th
the
one
measured
by
the
ARP
A
station.
The
a
v
erage
v
alue
of
carbon
monoxide
acquired
by
the
SmartBik
e
de
vice
and
calculated
o
v
er
180
samples
(1
e
v
ery
12
seconds)
is
1.2
0.5
ppm
(part
per
million).
Instead
the
one
monitored
by
the
ARP
A
station
1
,
calculated
as
the
a
v
erage
v
alue
of
the
tw
o
measures
acquired
at
14:00
and
15:00,
is
0.892
ppm
(1.1
mg
m
3
)
and
the
tw
o
measurements
are
compatible.
7.
CONCLUSION
AND
FUTURE
W
ORKS
In
this
paper
we
present
an
IoT
Cro
wd
Sensing
platform
that
of
fers
a
set
of
services
to
citizens
by
e
xploiting
a
netw
ork
of
bic
ycles
as
IoT
probes.
A
surv
e
y
aimed
at
identifying
the
most
interesting
bik
e-
enabled
services
for
users
w
as
conducted
among
288
users
that
usually
use
a
bik
e
in
their
daily
life.
The
follo
wing
services
were
identified:
a)
real
t
ime
remote
geo-location
detection
of
the
users’
bik
es,
b)
anti-theft
service,
c)
information
about
tra
v
eled
route
(distance,
duration,
and
rise),
and
d)
air
pollution
monitoring.
Then,
starting
from
an
enabling
s
cenario,
the
details
of
each
service
were
defined
and
the
architecture
of
the
SmartBik
e
platform
w
as
designed.
It
is
composed
of
three
main
components:
the
SmartBik
e
de
vices
for
data
collection,
the
end-user
de
vices
(e.g.,
smartphones
and
tablets)
as
user
interf
ac
es
for
the
real
time
bik
e
geo-
location
detection
and
the
anti-theft
service,
and
the
SmartBik
e
central
serv
ers
for
storing
re
v
ealed
data
and
pro
viding
a
web
interf
ace
for
data
visualization.
T
o
e
v
aluate
the
feasibility
of
the
approach
and
the
suitability
of
the
platform,
an
initial
prototype
of
the
presented
platform
w
as
implemented
and
the
platform
w
as
tested
by
a
v
olunteer
.
Results
demonstrate
that
the
proposed
SmartBik
e
platform
is
able
to
pro
vide
the
designed
services,
1
Actual
v
alues
acquired
through
the
ARP
A
monitoring
station
were
tak
en
from
the
ARP
A
of
ficial
website
[30].
IJECE
V
ol.
7,
No.
6,
December
2017:
3602
–
3612
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISSN:
2088-8708
3611
and,
in
addition,
that
the
accurac
y
of
the
air
pollution
measures
pro
vided
by
the
proposed
platform
are
compat-
ible
with
the
ones
pro
vided
by
the
of
ficial
en
vironmental
monitoring
system
of
the
city
of
T
urin.
Future
w
orks
will
include
the
de
v
elopment
of
missing
functionalities
and
a
lar
ge-scale
field
e
xperi
-
mentation:
the
described
platform
will
be
adopted
in
t
h
e
OpenAgor
`
a
project
[4]
within
the
T
orino
Li
ving
Lab
Campidoglio
e
xperimentation
[5].
The
OpenAgor
`
a
project
aims
at
helping
people
making
their
mobility
be-
ha
vior
more
sustainable
and,
at
the
same
time,
pro
viding
data
and
tools
that
can
be
used
by
the
city
institutions
to
enhance
the
o
v
erall
quality
of
life
of
their
citizens.
In
addition,
ne
w
air
quality
parameters
will
be
monitored
by
introducing
ne
w
sensors
in
the
Smart
Bik
e
de
vices.
Finally
,
starting
from
the
results
of
the
first
part
of
the
surv
e
y
re
g
arding
the
preferred
type
of
bic
ycle
and
the
most
used
bik
e
accessories,
the
aesthetic
design
of
the
SmartBik
e
de
vices
will
be
subjected
to
a
ne
w
study
that
will
in
v
olv
e
ag
ain
users.
A
CKNO
WLEDGMENT
This
w
ork
w
as
supported
by
a
fello
wship
from
TIM
(SW
ARM
Joint
Open
Lab)
for
T
eodoro
Mon-
tanaro.
The
authors
wish
to
thank
Roberta
Giannantonio
and
Ilaria
Zonda
who
designed
and
conducted
the
surv
e
y
about
SmartBik
e
features.
REFERENCES
[1]
A.
Cenedese,
A.
Zanella,
L.
V
angelista,
and
M.
Zorzi,
“P
ado
v
a
smart
city:
An
urban
internet
of
things
e
xperimentation,
”
in
W
orld
of
W
ir
eless,
Mobile
and
Multimedia
Networks
(W
oWMoM),
2014
IEEE
15th
International
Symposium
on
a
,
June
2014,
pp.
1–6.
[2]
A.
Zanella,
N.
Bui,
A.
Castellani,
L.
V
angelista,
and
M.
Zorzi,
“Internet
of
things
for
smart
cities,
”
IEEE
Internet
of
Things
J
ournal
,
v
ol.
1,
no.
1,
pp.
22–32,
Feb
2014.
[3]
L.
G.
Jaimes,
I.
J.
V
er
g
ara-Laurens,
and
A.
R
aij,
“
A
surv
e
y
of
incenti
v
e
techniques
for
mobile
cro
wd
sensing,
”
IEEE
Internet
of
Things
J
ournal
,
v
ol.
2,
no.
5,
pp.
370–380,
Oct
2015.
[4]
“OpenAgor
´
a
Project:
Mobility
as
a
service.
”
[Online].
A
v
ailable:
http://openagora.it/inde
x-en.html
[5]
“T
orino
Li
ving
Lab
-
Campidoglio.
”
[Online].
A
v
ailable:
http://torinoli
vinglab
.it/bandi/tllcampidoglio/
[6]
“Mo
v
e
Plus:
car
pooling,
”
http://www
.mo
v
eplus.it/
and
http://www
.easymoo
v
e.it/.
[7]
“Pon
yzero:
sustainable
urban
logistics.
”
[Online].
A
v
ailable:
http://www
.pon
yzero.com/
[8]
X.
Liu,
B.
Li,
A.
Jiang,
S.
Qi,
C.
Xiang,
and
N.
Xu,
“
A
bic
ycle-borne
sensor
for
monitoring
air
pollution
near
roadw
ays,
”
in
Consumer
Electr
onics
-
T
aiwan
(ICCE-TW),
2015
IEEE
International
Confer
ence
on
,
June
2015,
pp.
166–167.
[9]
F
.
Zeiger
and
M.
Huber
,
“Demonstration
abstract:
P
articipatory
sensing
enabled
en
vironmental
monitor
-
ing
in
smart
cities,
”
in
Pr
oceedings
of
the
13th
International
Symposium
on
Information
Pr
ocessing
in
Sensor
Networks
,
ser
.
IPSN
’14.
Piscata
w
ay
,
NJ,
USA:
IEEE
Press,
2014,
pp.
337–338.
[10]
X.
Liu,
C.
Xiang,
B.
Li,
and
A.
Jiang,
“Collaborati
v
e
bic
ycle
sensing
for
air
pollution
on
roadw
ay
,
”
in
2015
IEEE
12th
Intl
Conf
on
Ubiquitous
Intellig
ence
and
Computing
and
2015
IEEE
12th
Intl
Conf
on
A
utonomic
and
T
rusted
Computing
and
2015
IEEE
15th
Intl
Conf
on
Scalable
Computing
and
Communi-
cations
and
Its
Associated
W
orkshops
(UIC-A
TC-ScalCom)
,
Aug
2015,
pp.
316–319.
[11]
C.
V
agnoli,
F
.
Martel
li,
T
.
D.
Filippis,
S.
D.
Lonardo,
B.
Gioli,
G.
Gualtieri,
A.
Matese,
L.
Rocchi,
P
.
T
oscano,
and
A.
Zaldei,
“The
sensorwebbik
e
for
air
quality
monitoring
in
a
smart
city
,
”
in
Futur
e
Intellig
ent
Cities,
IET
Confer
ence
on
,
Dec
2014,
pp.
1–4.
[12]
Y
.
T
aniguchi,
K.
Nishii,
and
H.
Hisamatsu,
“Ev
aluation
of
a
bic
ycle-mounted
ultrasonic
distance
sen-
sor
for
monitoring
road
surf
ace
condition,
”
in
Computational
Intellig
ence
,
Communication
Systems
and
Networks
(CICSyN),
2015
7th
International
Confer
ence
on
,
June
2015,
pp.
31–34.
[13]
I.
Alam,
“
An
e
xploratory
in
v
estig
ation
of
user
in
v
olv
ement
in
ne
w
service
de
v
elopment,
”
J
ournal
of
the
Academy
of
Mark
eting
Science
,
v
ol.
30,
no.
3,
pp.
250–261,
2002.
[14]
C.-Y
.
Hsieh,
“Model
st
u
dy
for
temper
ature
microchange
by
wsn
technology
,
”
International
J
ournal
of
Electrical
and
Computer
Engineering
(IJECE)
,
v
ol.
2,
no.
5,
pp.
632–638,
October
2012.
[15]
S.
S.
P
.
Deekla,
R.
Phatthanakun
and
N.
Chomna
w
ang,
“
Al
microheater
and
ni
temperature
sensor
set
based-on
photolithograph
y
with
closed-loop
control,
”
International
J
ournal
of
Electrical
and
Computer
Engineering
(IJECE)
,
v
ol.
5,
no.
4,
pp.
849–858,
August
2015.
[16]
C.
Ryu
and
C.
Hur
,
“
A
monitoring
system
for
inte
grated
management
of
iot-based
home
netw
ork,
”
Inter
-
national
J
ournal
of
Electrical
and
Computer
Engineering
(IJECE)
,
v
ol.
6,
no.
1,
pp.
375–380,
February
SmartBik
e:
an
IoT
Cr
owd
Sensing
Platform
for
Monitoring
City
Air
P
ollution
(T
eodor
o
Montanar
o)
Evaluation Warning : The document was created with Spire.PDF for Python.