Ci a ion: Cal o, I.; Espin, A.;
Gil-Ga cía, J.M.; Fe nández
Bus aman e, P.; Ba ambones, O.;
Apiñaniz, E. Scalable IoT
A chi ec u e o Moni o ing IEQ
Condi ions in Public and P i a e
Buildings. Ene gies 2022,15, 2270.
h ps://doi.o g/10.3390/en15062270
Academic Edi o s: Nicu Bizon,
Mihai Op oescu, Philippe Pou e,
Rocío Pé ez de P ado and
Abdessa a Abdelke i
Recei ed: 4 Ma ch 2022
Accep ed: 17 Ma ch 2022
Published: 21 Ma ch 2022
Publishe ’s No e: MDPI s ays neu al
wi h ega d o ju isdic ional claims in
published maps and ins i u ional a il-
ia ions.
Copy igh : © 2022 by he au ho s.
Licensee MDPI, Basel, Swi ze land.
This a icle is an open access a icle
dis ibu ed unde he e ms and
condi ions o he C ea i e Commons
A ibu ion (CC BY) license (h ps://
c ea i ecommons.o g/licenses/by/
4.0/).
ene gies
A icle
Scalable IoT A chi ec u e o Moni o ing IEQ Condi ions in
Public and P i a e Buildings
Isid o Cal o 1,* , Ai ana Espin 1,*, Jose Miguel Gil-Ga cía2, Pablo Fe nández Bus aman e 3,
Osca Ba ambones 1,* and Es ibaliz Apiñaniz 4
1Sys em Enginee ing and Au oma ion Depa men , Facul y o Enginee ing o Vi o ia-Gas eiz,
Basque Coun y Uni e si y (UPV/EHU), 01006 Vi o ia-Gas eiz, Spain
2Depa men o Elec onic Technology, Facul y o Enginee ing o Vi o ia-Gas eiz,
Basque Coun y Uni e si y (UPV/EHU), 01006 Vi o ia-Gas eiz, Spain; [email p o ec ed]
3Depa men o Elec ical Enginee ing, Facul y o Enginee ing o Vi o ia-Gas eiz,
Basque Coun y Uni e si y (UPV/EHU), 01006 Vi o ia-Gas eiz, Spain; [email p o ec ed]
4Depa men o Applied Physics I, Facul y o Enginee ing o Vi o ia-Gas eiz,
Basque Coun y Uni e si y (UPV/EHU), 01006 Vi o ia-Gas eiz, Spain; [email p o ec ed]
*Co espondence: [email p o ec ed] (I.C.); [email p o ec ed] (A.E.);
osca [email p o ec ed] (O.B.)
Abs ac :
This pape p esen s a scalable IoT a chi ec u e based on he edge– og–cloud pa adigm
o moni o ing he Indoo En i onmen al Quali y (IEQ) pa ame e s in public buildings. Nowadays,
IEQ moni o ing sys ems a e becoming impo an o se e al easons: (1) o ensu e ha empe a u e
and humidi y condi ions a e adequa e, imp o ing he com o and p oduc i i y o he occupan s;
(2) o in oduce ac ions o educe ene gy consump ion, con ibu ing o achie ing he Sus ainable
De elopmen Goals (SDG); and (3) o gua an ee he quali y o he ai —a key conce n due o he
COVID-19 wo ldwide pandemic. Two kinds o nodes compose he p oposed a chi ec u e; hese
a e he so-called: (1) sma IEQ senso nodes, esponsible o acqui ing indoo en i onmen al
measu es locally, and (2) he IEQ concen a o s, esponsible o collec ing he da a om sma senso
nodes dis ibu ed along he acili ies. The IEQ concen a o s a e also esponsible o con igu ing
he acquisi ion sys em locally, logging he acqui ed local da a, analyzing he in o ma ion, and
connec ing o cloud applica ions. The p esen ed a chi ec u e has been designed using low-cos
open-sou ce ha dwa e and so wa e—speci ically, single boa d compu e s and mic ocon olle s
such as Raspbe y Pis and A duino boa ds. WiFi and TCP/IP communica ion echnologies we e
selec ed, since hey a e ypically a ailable in co po a i e buildings, bene i ing om al eady a ailable
communica ion in as uc u es. The applica ion laye was implemen ed wi h MQTT. A p o o ype was
buil and deployed a he Facul y o Enginee ing o Vi o ia-Gas eiz, Uni e si y o he Basque Coun y
(UPV/EHU), using he exis ing ne wo k in as uc u e. This p o o ype allowed o collec ing da a
wi hin di e en academic scena ios. Finally, a sma senso node was designed including low-cos
senso s o measu e empe a u e, humidi y, eCO2, and VOC.
Keywo ds: IoT; WSN; IEQ; IAQ; SDGs; MQTT; Raspbe y Pi; A duino; open sou ce
1. In oduc ion
Indus y 4.0 b ings shi s based on he in oduc ion o mode n echnologies o in-
c easing in e connec i i y and sma au oma ion [
1
]. Decen alized communica ion in as-
uc u es, mainly hose ha a e wi eless, a e inc easingly adop ed in hese applica ions [
2
].
In his scena io, he In e ne o Things (IoT) makes i possible o connec objec s o hings
o he In e ne , wi h he pu pose o collec ing da a and con olling p ocesses o machines
emo ely, by means o mesh s uc u es ha allow ubiqui ous communica ions [
3
]. Indeed,
he numbe o connec ed objec s is inc easing exponen ially, eaching 20 billion connec ed
“ hings” by 2020 [
4
]. These objec s in e ac wi h each o he , coope a ing o achie e a goal.
Ene gies 2022,15, 2270. h ps://doi.o g/10.3390/en15062270 h ps://www.mdpi.com/jou nal/ene gies
Ene gies 2022,15, 2270 2 o 23
They a e capable o gene a ing la ge amoun s o da a, and i is necessa y o c ea e so wa e
ha is capable o collec ing he in o ma ion om di e en loca ions o analyze hem. The
da a a e g ows abou 40% each yea [
4
], and i is no only he olume o da a ha in-
c eases, bu also he speed and a ie y. Consequen ly, mode n applica ions mus deal wi h
la ge amoun s o da a, which b ings compu a ional challenges o da a s o age, analysis,
and isualiza ion.
IoT sys ems a e widely sp ead in e y di e en ields. Howe e , hey ha e o mee
a ious equi emen s: (1) dealing wi h he e ogenei y, since di e en pla o ms a e in-
ol ed; (2) using esou ce-cons ained de ices, such as sma senso s; (3) applica ions
ha equi e spon aneous in e ac ion; (4) ul a-la ge-scale ne wo ks and a la ge numbe o
e en s;
(5) dynamic
ne wo k beha io equi emen s; (6) con ex -awa e and loca ion-awa e
applica ions; and (7) he need o dis ibu ed in elligence [5].
Cu en ly, sus ainabili y is gaining inc easing impo ance and, consequen ly, socie y is
becoming conce ned abou i . Sus ainable de elopmen is based on h ee pilla s: economic
sus ainabili y, en i onmen al sus ainabili y, and social sus ainabili y. In Sep embe 2015,
he 2030 UN Agenda o Sus ainable De elopmen was p oposed, de ining 17 Sus ainable
De elopmen Goals (SDGs) ha in e link he h ee aspec s o sus ainable de elopmen ha
we e men ioned be o e [6].
The Uni e si y o he Basque Coun y (UPV/EHU) launched an in e nal p og am, he
Campus Bizia Lab (CBL), aimed a in oducing he SDGs in he uni e si y. The p esen
wo k, which is pa o his ini ia i e, add esses h ee speci ic SDGs: sus ainable ci ies and
communi ies, esponsible consump ion and p oduc ion, and good heal h and well-being.
Mo e speci ically, his wo k aims a moni o ing he en i onmen al condi ions a he
uni e si y acili ies o imp o e he sus ainabili y o he buildings and well-being o he
occupan s. The p oblem ha igge ed his wo k was he de ec ion o se e al he mal
discom o si ua ions a he uni e si y acili ies: On he one hand, some places we e
cold al hough he cen al hea ing sys em was on, equi ing auxilia y elec ic adia o s,
which a e less e icien . On he o he hand, o he loca ions we e o e -hea ed, so he e was
a consequen
was e o ene gy as i was necessa y o open he windows, which ob iously
has a bad e ec on sus ainabili y.
This is a common issue as nowadays, buildings u ilize a ound 40% o global en-
e gy consump ion and, consequen ly, he e is a need o making hese buildings mo e
ene gy-e icien [7]
. A combina ion o se e al app oaches may help o de ec and sol e
his p oblem, including be e isola ing ma e ials, use o low-powe appliances, and de-
ploymen o ene gy managemen sys ems. In his scena io, Indoo En i onmen Quali y
(IEQ) moni o ing sys ems a e necessa y o de ec ene gy-was e si ua ions and con ibu e o
de eloping solu ions. Since his is a wo ldwide p oblem, i is impo an o achie e e sa ile
and low-cos IEQ moni o ing sys ems ha a e easy o implan in di e en buildings.
The p oblem o ene gy was e, ypically ound in o ices and public buildings, was
p esen ed o he s uden s o he B.Sc. Deg ee in Indus ial Elec onics and Au oma ion
Enginee ing o aising awa eness abou he SDGs, encou aging hem o conside how
hey could con ibu e as u u e enginee s o c ea ing a mo e sus ainable wo ld by means o
conc e e ac i i ies. In pa icula , s uden s de eloped a p elimina y IEQ moni o ing sys em
based on IoT, which he ins uc o s had simpli ied o jus measu e empe a u e alues. This
p ojec was p oposed as a P ojec Based Lea ning ac i i y o he s uden s [
8
]. These kinds
o ini ia i es in ol e s uden s om a pe spec i e o lea ning h ough p ac ice, making he
en i e scien i ic p ocess mo e inclusi e. Fo example, And eo i e al. [
9
] p esen a me e ing
ho box o analyzing insula ion echnologies a his o ic buildings, which was ini ially
de eloped as an educa ional ac i i y.
This wo k goes beyond, since i desc ibes a ull IoT scalable a chi ec u e aimed a
moni o ing indoo en i onmen al quali y inside public buildings, such as uni e si y ac-
ul ies, which uns o e exis ing communica ion in as uc u es. The p oposed sys em
allows: (1) ensu ing ha empe a u e and humidi y condi ions a e adequa e, imp o ing he
com o and p oduc i i y o he occupan s; (2) in oducing ac ions o a oid ene gy was ing,
Ene gies 2022,15, 2270 3 o 23
con ibu ing o achie e he SDG; and (3) gua an eeing he quali y o he ai , a key conce n
wi h he COVID-19 wo ldwide pandemic. The p oposed IEQ moni o ing sys em, which
may be adap ed o moni o di e en en i onmen al magni udes, uses low-cos open-sou ce
ha dwa e and so wa e. A p o o ype o he p oposed sys em was deployed a he Facul y
o Enginee ing o Vi o ia-Gas eiz o moni o he Indoo Ai Quali y (IAQ) by means o
di e en low-cos senso s.
The p esen ed sys em akes empe a u e and ela i e humidi y measu emen s. These
magni udes allow de ec ing uncom o able si ua ions o loca ions in which ene gy is was ed.
La e , i may be used o implemen co ec i e ac ions aimed a gua an eeing he com o o
he Uni e si y communi y while egula ing he ene gy consump ion, con ibu ing o he
sus ainabili y o he buildings.
In addi ion, he quali y o he ai inside he acili ies was moni o ed, as IAQ is di ec ly
ela ed o human heal h. Ac ually, i has been ound ha exposu e o high concen a ions
o ai pollu an s causes nega i e consequences in humans, om mild heal h issues such
as
a dec ease
in cogni i e abili ies and p oduc i i y [
10
], o mo e se ious issues, includ-
ing espi a o y and ca dio ascula illness, alle gic symp oms, cance s, and p ema u e
mo ali y [11]
. Mo eo e , due o he COVID-19 pandemic, ai quali y and cons an en ila-
ion a e gaining impo ance in ela ion o p e en ing he sp ead o he i us in closed a eas.
Fo his aim, eCO
2
and VOC we e moni o ed. These measu emen s could help o decide
when o ake ac ions such as opening o closing he windows o inc ease en ila ion [12].
The p oposed moni o ing sys em is based on he edge– og–cloud pa adigm [
13
]. This
app oach p oduces scalable and easily con igu able a chi ec u es. This kind o a chi ec u es
dis ibu e he in elligence, compu a ion, and s o age in loca ions close o he sou ce o da a,
bu also p o ides connec i i y o cloud se ices. This app oach eases i s deploymen a
uni e si y acili ies o buildings wi h simila cha ac e is ics, such as o ice buildings o
indus ial acili ies. These a e places whe e people spend a g ea amoun o hei ime, so i
is e y impo an o ha e good en i onmen al condi ions o imp o e hei li e quali y.
The p oposed sys em was deployed using he a ailable communica ions in as uc u e
o he ins i u ion (UPV/EHU), a oiding he ins alla ion o ad hoc de ices. Namely, he
co po a i e
WiFi is al eady deployed all o e he acili ies, co e ing almos e e y co ne
o he buildings. In o de o sepa a e he a ic o igina ed by he IEQ moni o ing sys em
om he mains eam ne wo k a ic, caused by academic ac i i y, a VLAN (Vi ual Local
A ea Ne wo k) was used. A he applica ion laye , he MQTT p o ocol was used due o i s
simplici y and he ac ha his is one o he mos popula p o ocols used a IoT sys ems.
Low-cos open-sou ce ha dwa e was used; A duino MKR WIFI 1010 boa ds powe ed
wi h ba e ies we e used as sma IEQ senso nodes and a Raspbe y Pi 3B+ was used as
an IEQ
Concen a o o collec ing he da a om all nodes. A PCB (P in ed Ci cui Boa d)
was designed speci ically o hold he selec ed low-cos senso s and A duino boa ds. The
achie ed sys em is scalable, making i possible o ha e di e en amoun s o nodes, wi h
each node holding se e al senso s o measu ing a b oad numbe o a iables. I is also
emo e-con igu able since i is possible o change he con igu a ion o he whole sys em
di ec ly om he cen al node.
The layou o he a icle is as ollows. Sec ion 2summa izes ela ed wo k. Sec ion 3
desc ibes in de ail he design o he p oposed a chi ec u e. Sec ion 4p esen s he alida ion
o he p o o ype deployed a he Facul y o Enginee ing o Vi o ia-Gas eiz and some
measu emen s aken while in he alida ion p ocess. Finally, some conclusions a e d awn.
2. Rela ed wo k
2.1. IEQ Moni o ing Sys ems
Du ing he las yea s, se e al wo ks ha e p esen ed di e en app oaches aimed a
e alua ing he com o in bo h esiden ial and comme cial buildings. The wo k in [
14
]
p esen s a c i ical e iew o s udies and in es iga ed occupan com o by means o en i-
onmen al and non-en i onmen al a iables. Some s udies collec subjec i e da a ela ed o
how occupan s pe cei e indoo en i onmen s, which is ypically acqui ed ia occupan sa -
Ene gies 2022,15, 2270 4 o 23
is ac ion su eys [15]. Howe e , mos s udies a e based on IEQ pa ame e s equi ing IEQ
moni o ing sys ems, which measu e physical en i onmen al changes ha can be quan i ied
using di e se equipmen .
Typically, IEQ consis s o ou majo a iables: ai quali y, he mal com o , isual
com o , and acous ic com o , and many di e en pa ame e s ha e e ec s on each o
hem [
10
]. Fo ai quali y, he mos measu ed pa ame e s h ough he published s udies a e
CO
2
, CO, PM10, PM2.5, and VOC [
16
]. The mos common pa ame e s o he mal com o
a e empe a u e, ela i e humidi y, and ai cu en s, bu clo hing and physical ac i i y also
ha e g ea impac . F equen ly, isual com o is moni o ed by luminosi y o ligh in ensi y,
whe eas acous ic com o is e alua ed by checking he noise le el [10].
Some s anda ds such as ASHRAE 55 [
17
], RESET Ai [
18
], and ISO 7730 [
19
] de ine
he app op ia e ange o alues o hese pa ame e s. Table 1summa izes hose ha apply
o his wo k.
Table 1. Accep ed ange o alues o en i onmen al pa ame e s acco ding o di e en s anda ds.
Va iable App op ia e Values S anda d
CO2Less han 1000 ppm ASHRAE
VOC Less han 250 ppb RESET Ai
Tempe a u e (1.2 m) Be ween 23.3 and 27.8 ◦C ASHRAE
Rela i e humidi y Less han 65% ASHRAE
The moni o ing o hese pa ame e s, especially CO
2
and empe a u e, allows o
adequa ely en ila ing buildings, o example, by means o au oma ic en ila ion sys ems
ha exchange low-quali y ai wi h esh ai , and his is key o educing he sp ead o
COVID-19 [
12
]. This p oblem is especially ele an in public buildings in win e since i is
necessa y o achie e a comp omise be ween he indoo s empe a u e and he quali y o
he ai .
In his scena io, IEQ moni o ing sys ems a e essen ial o main aining hese pa ame e s
inside he accep ed anges. Mo eo e , his kind o sys em leads he way o implemen ing
con ol ac ions such as con olling he hea ing sys em, en ila ing he oom, o pu i ying
he ai when necessa y.
Much ecen esea ch abou indoo en i onmen al moni o ing sys ems can be ound
in he li e a u e. Howe e , o da e, all wi eless gas senso ne wo ks ace a ade-o be ween
senso node cos and da a quali y because no sui able, low-cos echnology o speci ic,
quan i a i e chemical analysis is cu en ly a ailable [
20
]. Ac ually, he main obs acle is he
cu en lack in sui able chemical analysis echnologies o de e mine he concen a ion o
gaseous ai pollu an s speci ically and sensi i ely a low-cos , o e ing long- e m, s able
de ec ion [
21
]. Some wo ks explo e he po en ial o elec onic noses ha make use o
commodi y gas senso s based on MOS and MEMS echnologies. Fo example, TheOdo
consis s o wo closed measu ing boxes wi h senso s, each connec ed o and con olled by
a Raspbe y Pi [22].
The wo k in [
23
] discusses he possibili ies o IoT sys ems o measu ing some gases,
as well as he majo echnologies a ailable. This a icle summa izes di e en ypes o gas-
senso s and communica ion echnologies. Fo example, [
24
] p esen s a sys em ha in o ms
oom occupan s abou bad ai quali y wi h ambien ligh s so ha hey ake co ec i e
ac ions. In his sys em, IEQ measu es a e aken wi h senso s a ached o a Raspbe y Pi. The
wo k in [
25
] p esen s an indoo en i onmen al sys em a chi ec u e wi h cloud connec i i y
ha akes en i onmen al measu emen s in public buildings by means o senso s di ec ly
a ached o ZigBee nodes. Ano he IEQ moni o ing sys em is p esen ed in [
21
]. Thei
app oach uses PSoC mic ocon olle s and Z-Wa e communica ion echnology. This sys em
was es ed in i e ooms o a school. Simila ly, a wi eless senso ne wo k, based on XBee
echnology, is p oposed in [
26
] o moni o ing ai quali y in a lib a y. In [
27
], a wi eless
senso ne wo k is used o measu ing en i onmen al da a in open a eas by posi ioning
he nodes in di e en poin s o he ci y. Some wo ks ecommend he in oduc ion o
Ene gies 2022,15, 2270 5 o 23
low-cos senso s o educing he accumula ion o CO
2
in indoo en i onmen s in selec ed
loca ions [
28
]. The ac ual wo ldwide COVID-19 pandemic has also led o indoo ai quali y
moni o ing s udies like he one p oposed in [11].
Howe e , IEQ moni o ing is no only use ul o augmen ing he com o and heal h
o he building occupan s, bu also o ene gy sa ing. By moni o ing he en i onmen al
pa ame e s o each oom, i is possible o mo e om a cen alized con ol o he hea ing
sys em o indi idualized con ol o a oom. This way, si ua ions in which windows a e
open while he hea ing sys em is on can be a oided, sa ing huge quan i ies o ene gy while
main aining good en i onmen al condi ions and imp o ing he sus ainabili y objec i es by
aking conc e e ac ions. In his di ec ion, a wi eless senso ne wo k o con olling he ai
condi ioning sys em is p oposed in [
29
]. The ene gy pe o mance along wi h he indoo
en i onmen al condi ions o a uni e si y campus do mi o y a e analyzed in [
30
] by means
o indoo en i onmen al quali y measu emen s. Also, Ma in-Ga ín e al. [
31
] p esen
a low-cos
building moni o ing sys em o di e en IEQ a iables based on open-sou ce
pla o ms. In his case, da a a e s o ed on a lash memo y ca d, bu he au ho s do no
p o ide much in o ma ion abou he communica ion echnologies used; hey only men ion
he use o WiFi echnology o connec wi h he cloud.
A duino-like boa ds ha e p o en o be capable o being used as sma senso s. Fo
example, [
31
–
33
] desc ibe wo se ups o IEQ moni o ing based on A duino and XBee
echnology. Also, in [
34
], an A duino-based sys em is p esen ed o he measu emen o
se e al pa ame e s in ol ed in wa e quali y. Some wo ks use al e na i es, o example he
design p esen ed in [
31
] is based on he ESP-8266 boa d, whe eas an ad hoc PCB based on
an Mic ochip ATmega328P is p esen ed in [35].
The moni o ing sys em p esen ed in his pape is aimed a sol ing se e al o hese
issues, namely, educing ene gy consump ion, imp o ing he com o o he building
occupan s, and ensu ing a heal hy quali y o ai ha can p e en s uden s and s a om
becoming in ec ed wi h COVID-19.
2.2. Communica ion Technologies o IoT Applica ions
The numbe o connec ed de ices is exponen ially inc easing wi hin Indus y 4.0
and In e ne o Things (IoT) pa adigms. These de ices gene a e huge quan i ies o da a
ha mus be e icien ly in eg a ed by means o so wa e applica ions. Mo eo e , hese
so wa e applica ions mus allow o he dis ibu ed collec ion o he measu emen s and
be in e ope able and scalable o in oduce lexibili y and eusabili y, in o de o be used in
di e en scena ios.
Some eme ging echnologies ha e been a ge ed o implemen ing cloud and og com-
pu ing IoT applica ions [
36
]. Recen ly, he in oduc ion o hese echnologies and low-cos
open-sou ce ha dwa e and so wa e has allowed o he c ea ion o a o dable wi eless
moni o ing sys ems. Since a dis ibu ed moni o ing sys em equi es a la ge amoun o
nodes, i is impo an ha he cos o each node, as well as he ene gy consump ion, is
educed. Many wo ks based on low-cos de ices can be ound in he li e a u e. Typi-
cally, hey include single-boa d compu e s (SBC), such as Raspbe y Pi, o low-cos sma
senso s [22,24,37]
. Some wo ks, such as [
38
,
39
], a e aimed a building low-cos sys ems
o moni o ing and e en ually con olling mic og id sys ems. O he sys ems allow mea-
su ing ele an me eo ological a iables and acqui ing pho o ol aic da a di ec ly om
he plan s [
40
]. Simila app oaches can be ound o assessing hyg o he mal condi ions in
his o ic buildings by means o di e en communica ion echnologies [9,41].
Rega ding communica ion echnologies o Wi eless Senso Ne wo ks (WSN), he
mos ypical echnologies a e Blue oo h, ZigBee, and WiFi. Blue oo h Low Ene gy (BLE)
was c ea ed o expanding he use o Blue oo h o IoT applica ions, as i consumes less
ene gy and is able o wo k wi h ba e y-powe ed elemen s. Many moni o ing sys ems use
his echnology [
42
–
44
]. Howe e , BLE equi es he use o ga eways o he ansmission
o he da a in la ge ne wo ks, since i has a limi ed co e age. ZigBee also consumes
e y li le ene gy and, consequen ly, i is a echnology equen ly used in moni o ing
Ene gies 2022,15, 2270 6 o 23
sys ems [
25
,
26
,
45
,
46
], bu i equi es addi ional ga eways o ecollec ing he da a and
sending i o he In e ne . In addi ion, hey equi e o ming a mesh ne wo k wi h ou e
nodes o co e la ge a eas. O he wo ks use Z-Wa e echnology, o example, [
21
] p esen s
a sma node o sensing empe a u e, humidi y, and CO
2
concen a ion. Some app oaches
use a combina ion o se e al echnologies, as in [
47
], which combines ZigBee, WiFi, and
MQTT o building IoT applica ions.
Since WiFi in as uc u e is al eady deployed in mos co po a i e buildings, i is
an in e es ing
al e na i e o hese kinds o sys ems. E en hough i consumes mo e ene gy
han o he op ions, i does no equi e any ex a in e sion and is he e o e a low-cos
in as uc u e. Fo his eason, many s udies use his echnology [48–50].
Se e al messaging applica ion p o ocols aimed a IoT applica ions exis , and
a comp ehensi e
e iew is p esen ed in [
51
]. The mos common a e MQTT (Message Queu-
ing Teleme y T anspo ), CoAP (Cons ained Applica ion P o ocol), AMQP (Ad anced
Message Queuing P o ocol), HTTP (Hype Tex T ans e P o ocol), XMPP (Ex ensible Mes-
saging and P esence P o ocol), and DDS (Da a Dis ibu ion Se ice). MQTT and CoAP we e
speci ically c ea ed o da a ecollec ing applica ions. AMQP was c ea ed o quick and
secu e comme cial ansac ions and HTTP o web applica ions ha communica e h ough
he In e ne [
52
]. XMPP was c ea ed o sol ing he e ogenei y p oblems on IoT applica ions
and DDS o eal- ime messages [
53
]. XMPP and DDS a e used o high pe o mance
applica ions ha equi e wi h e y s ic QoS equi emen s, such as sma g ids [
54
] o
ac o y au oma ion [
55
], and hey a e mo e complex o use han o he al e na i es. Finally,
AMQP and HTTP we e no c ea ed o IoT applica ions, and hey use bigge message sizes,
equi ing mo e ene gy consump ion and bigge la ency han MQTT and CoAP [52].
MQTT and CoAP ha e simila cha ac e is ics in e ms o message size and o e head,
ene gy consump ion and esou ce equi emen s, and bandwid h and la ency. In hese
espec s, CoAP is somewha ligh e han MQTT due o he ac ha i uses UDP ins ead o
TCP. Howe e , MQTT is mo e eliable han CoAP, and his ac makes i mo e adequa e.
Finally, MQTT is also one o he simples p o ocols a ailable o c ea ing his kind o
applica ion. Fo hese easons, MQTT has become he mos popula and widesp ead IoT
p o ocol o Machine- o-Machine (M2M) applica ions [
56
]. Ac ually, mos IEQ moni o ing
sys ems ound in he li e a u e end o use MQTT [37,49,57–59].
Se e al emo e IEQ moni o ing comme cial sys ems can be ound on he
ma ke [60–62]
.
Howe e , he p oposed solu ion is an open solu ion based on ee open-sou ce ha dwa e
and so wa e. Thus, he sou ce code may be easily adap ed o al e na i e scena ios in-
ol ing di e en measu ed alues [
63
]. The selec ion o he echnologies ensu es high
in e ope abili y and scalabili y. Ac ually, new magni udes could be easily added using
di e en ypes o senso s: analogue, digi al (by means o he SPI o I2C senso buses).
Besides, he exis ing uni e si y WiFi in as uc u e was used h ough a p i a e VLAN and
MQTT communica ions.
3. Scalable IoT A chi ec u e o IEQ Sys ems
This sec ion desc ibes in de ail an IoT a chi ec u e o moni o ing IEQ condi ions.
Namely, i desc ibes an a chi ec u e ha ollows he edge– og–cloud compu ing
pa adigm [
3
]. The hie a chical and collabo a i e edge– og–cloud a chi ec u es p o ide
emendous bene i s as hey enable he dis ibu ion o in elligence and compu a ion o
achie e an op imal solu ion while sa is ying he gi en cons ain s, e.g., delay ene gy
adeo [13]
. In edge and og pa adigms, compu ing and s o age esou ces a e loca ed nea
he sou ce o da a, bu hese de ices may also connec wi h cloud se ices. This app oach
p oduces di e se bene i s such as scalabili y, ubiqui y, eliabili y, and high-pe o mance,
among o he s.
The p oposed a chi ec u e adap s he edge– og–cloud pa adigm o be used o moni o -
ing IEQ condi ions in wide buildings whe e he e is WiFi in as uc u e al eady deployed,
which is qui e common in public o p i a e buildings.
Ene gies 2022,15, 2270 7 o 23
3.1. A chi ec u e O e iew
This a chi ec u e has wo ypes o nodes, he so-called sma IEQ senso nodes, which a e
esponsible o moni o ing he measu ed a iables nex o hei loca ion, and he so-called
IEQ concen a o s, which a e esponsible o logging and analyzing he da a acqui ed by
a g oup o sma IEQ senso nodes. The IEQ concen a o also p o ides connec i i y o
di e en cloud se ices such as da a s o age and analy ics, isualiza ion, o e en usage
p edic ions based on wea he o ecas .
Nowadays, hese nodes may be implemen ed by means o low-cos open-sou ce
ha dwa e such as single boa d compu e s as IEQ concen a o s, e.g., Raspbe y Pi, and
single boa d mic ocon olle s as sma IEQ senso nodes, e.g., A duino o ESP8266 boa ds.
Figu e 1depic s he implemen a ion o he p oposed h ee- ie a chi ec u e. A he
bo om, he edge laye , a e he Sma IEQ senso s nodes, which a e esponsible o acqui ing
local alues o he IEQ moni o ed a iables such as empe a u e, humidi y, concen a ion
o di e se componen s such as CO
2
o VOC componen s. In he middle, a he og laye ,
a e he IEQ Concen a o s, which a e esponsible o collec ing he alues om a g oup
o sma IEQ senso nodes loca ed in an a ea. These componen s a e also esponsible o
con igu ing he sma senso s in one a ea, as well as p o iding local s o age and basic da a
isualiza ion. IEQ Concen a o s a e connec ed o cloud se ices loca ed a he op laye
o di e en asks, including global ime, da a s o age and analy ics, isualiza ion, wea he
o ecas p edic ions aimed a educing ene gy consump ion, and so on.
Ene gies 2022, 15, x FOR PEER REVIEW 8 o 25
Figu e 1. O e all a chi ec u e o he wi eless acquisi ion sys em.
This p oposed a chi ec u e allows o implemen ing a se o unc ionali ies, which
a e desc ibed below:
• Cen alized con igu a ion: The con igu a ion o he whole IEQ moni o ing sys em is
cen alized a he cloud. Howe e , IEQ Concen a o s hold he local con igu a ion o
a se o sma IEQ senso nodes. This app oach in oduces highe lexibili y o
changing he con igu a ion o he sma IEQ senso nodes since i is no necessa y o
ep og am hem. In addi ion, i is possible o check he in eg i y o he IEQ moni o -
ing sys em be o e i s ope a ion. A powe ing-up ime, sma IEQ senso nodes send
a message ha includes hei iden i ie o he closes IEQ Concen a o , eques ing
con igu a ion. Upon ecep ion o he con igu a ion, sma IEQ senso nodes sel -con-
igu e acco dingly. The con igu a ion o e e y sma senso includes se e al pa am-
e e s including loca ion, quan i y o a ached senso s, and he cha ac e is ics o all
speci ic senso s connec ed.
• Va iable sampling imes: IEQ Concen a o s use a local clock o in e oga e IEQ sen-
so nodes acco dingly. This clock is synch onized o all de ices in he IEQ moni o -
ing sys em (cloud and og laye s) by means o he NTP p o ocol, execu ed a he
cloud. The p oposed app oach is adequa e o sampling imes in he ange o
minu es, which he au ho s conside adequa e o IEQ moni o ing sys ems since IEQ
a iables do no change ab up ly. The IEQ Concen a o s a e esponsible o sending
MQTT messages o he sma IEQ senso s, indica ing when hey ha e o ake he
measu es o he IEQ a iables. Thus, upon ecep ion o he co esponding MQTT
opic, he sma IEQ senso nodes ake new measu emen s om he a ached senso s
and send he cap u ed alues o he nex IEQ Concen a o by means o MQTT opics.
This app oach allows o synch onizing o he measu es aken a di e en a eas o
he building, e en when se e al IEQ Concen a o s a e in ope a ion. In addi ion, i is
possible o dynamically change he sampling ime while he sys em is in ope a ion,
i necessa y.
• Local collec ion o da a: The IEQ concen a o collec s he da a o on-line and o -
line analysis om he a ached sma IEQ senso nodes. These nodes hold he selec ed
senso s o measu ing he IEQ pa ame e s. Collec ed da a a e s o ed locally a he
IEQ concen a o , which is esponsible o sending hem pe iodically o he cloud,
Figu e 1. O e all a chi ec u e o he wi eless acquisi ion sys em.
Wi eless echnologies we e used o implemen he communica ion be ween he sma
IEQ senso s and he IEQ Concen a o . This app oach allows eaching almos e e y co ne
o buildings wi h WiFi co e age, which commonly eaches e e ywhe e in public and
p i a e buildings. The MQTT p o ocol was used on op o TCP/IP since his is a e sa ile
and easy- o-use al e na i e ypically used o c ea ing non-c i ical IoT applica ions. In
addi ion, he use o he MQTT p o ocol allows us o adequa ely scale he IEQ moni o ing
sys em o di e en layou s and sizes.
This p oposed a chi ec u e allows o implemen ing a se o unc ionali ies, which a e
desc ibed below:
Ene gies 2022,15, 2270 8 o 23
•Cen alized con igu a ion:
The con igu a ion o he whole IEQ moni o ing sys em is
cen alized a he cloud. Howe e , IEQ Concen a o s hold he local con igu a ion
o
a se
o sma IEQ senso nodes. This app oach in oduces highe lexibili y o
changing he con igu a ion o he sma IEQ senso nodes since i is no necessa y o
ep og am hem. In addi ion, i is possible o check he in eg i y o he IEQ moni o ing
sys em be o e i s ope a ion. A powe ing-up ime, sma IEQ senso nodes send
a message ha includes hei iden i ie o he closes IEQ Concen a o , eques ing
con igu a ion. Upon ecep ion o he con igu a ion, sma IEQ senso nodes sel -
con igu e acco dingly. The con igu a ion o e e y sma senso includes se e al
pa ame e s including loca ion, quan i y o a ached senso s, and he cha ac e is ics o
all speci ic senso s connec ed.
•Va iable sampling imes:
IEQ Concen a o s use a local clock o in e oga e IEQ sen-
so nodes acco dingly. This clock is synch onized o all de ices in he IEQ moni o ing
sys em (cloud and og laye s) by means o he NTP p o ocol, execu ed a he cloud.
The p oposed app oach is adequa e o sampling imes in he ange o minu es, which
he au ho s conside adequa e o IEQ moni o ing sys ems since IEQ a iables do no
change ab up ly. The IEQ Concen a o s a e esponsible o sending MQTT messages
o he sma IEQ senso s, indica ing when hey ha e o ake he measu es o he IEQ
a iables. Thus, upon ecep ion o he co esponding MQTT opic, he sma IEQ
senso nodes ake new measu emen s om he a ached senso s and send he cap u ed
alues o he nex IEQ Concen a o by means o MQTT opics. This app oach allows
o synch onizing o he measu es aken a di e en a eas o he building, e en when
se e al IEQ Concen a o s a e in ope a ion. In addi ion, i is possible o dynamically
change he sampling ime while he sys em is in ope a ion, i necessa y.
•Local collec ion o da a:
The IEQ concen a o collec s he da a o on-line and o -line
analysis om he a ached sma IEQ senso nodes. These nodes hold he selec ed
senso s o measu ing he IEQ pa ame e s. Collec ed da a a e s o ed locally a he
IEQ concen a o , which is esponsible o sending hem pe iodically o he cloud,
whe e hey a e s o ed and analyzed o di e en pu poses. IEQ concen a o s allow
bo h on-line and o -line basic analysis o he cap u ed da a, by means o an HMI
applica ion. This applica ion allows di e en ope a ions such as plo ing he da a o
selec ed senso s, he calcula ion o ypical alues such as maximum, minimum, and
mean alues and s anda d de ia ions, and he pe cen age o alid measu emen s o
each node and senso o a selec ed day o ime pe iod.
•Cloud se ices:
IEQ concen a o s ac as ga eways be ween og and cloud se ices,
by means o MQTT connec i i y. Typically, he adop ion o edge and og pa adigms
allow dealing wi h he massi e amoun s o aw da a o IoT applica ions. Cloud
se ices allow o scaling he IEQ moni o ing sys em o each e e y co ne o e y
la ge buildings o a eas, such as a whole campus, cen alizing all he in o ma ion.
This laye is composed by di e en se ices, e.g., he global ime o all de ices a he
sys em, da a s o age o all de ices, ad anced da a analy ics and isualiza ion, and
usage ecommenda ions o IEQ esou ces based on wea he o ecas and measu ed
pa ame e s, such as opening/closing he windows o be e en ila ion o swi ching
o he hea e s. Finally, hese se ices a e esponsible o sending ala m messages o
ope a o s i he measu ed alues a e ou o he speci ied bounds.
3.2. Communica ion Technologies
Since he sys em mus co e whole buildings, such as a acul y o e en se e al build-
ings on one campus, he dis ances among he sma IEQ senso nodes used o acqui ing he
moni o ed a iables may become qui e la ge. Fo ha eason, i would no be possible o
connec he de ices ia Blue oo h wi hou using ga eways. Using Zigbee, o a combina ion
o echnologies, as in [
47
], would p obably ha e also been adequa e, bu he inal decision
was o choose WiFi echnology. This al e na i e equi es a lowe in e sion since in public
and p i a e buildings, as in acul y acili ies, i is ypically deployed o each e e y co ne
Ene gies 2022,15, 2270 9 o 23
o he buildings. Mo eo e , i is ecommended o use a VLAN (Vi ual Local A ea Ne wo k)
in ane o sepa a e he IEQ moni o ing a ic om mains eam In e ne a ic. Thus, he
VLAN ne wo k allows IEQ concen a o s o collec he alues acqui ed by he sma IEQ
senso nodes. These nodes may use a di e en connec ion o access he cloud se ices
loca ed on he In e ne . In addi ion, he use o his VLAN imp o es he secu i y o he IEQ
moni o ing sys em as only es ic ed de ices may connec o he ne wo k.
As shown in Figu e 1, MQTT is used be ween he edge and og laye s o es ablishing
he connec i i y among he sma IEQ senso nodes and he IEQ concen a o s and also
be ween he og and cloud laye s, o connec he IEQ concen a o s wi h cloud se ices.
This p o ocol was chosen due o i s good pe o mance in non- ime-c i ical applica ions
(such as IEQ sys ems) wi h small size messages. Also, i was conside ed as easy o use in
IoT applica ions. Ac ually, MQTT is he mos common communica ion choice o building
M2M IoT applica ions [55].
The MQTT p o ocol uns o e TCP/IP and ollows he publish/subsc ibe pa adigm.
The use o he TCP p o ocol ensu es ha messages a i e co ec ly and in he same o de
ha hey we e sen . I p o ides connec i i y among se e al de ices by means o a MQTT
b oke . Some applica ions publish messages, deli e ed o he b oke by means o a opic.
The b oke is esponsible o sending he messages o all applica ions ha a e subsc ibed o
ha speci ic opic. Mo eo e , opics can be di ided in o sub opics and he subsc ip ion o
he en i e opic o only a sub opic is possible. I is con enien o use ixed IP add esses o
iden i y e e y de ice in he IEQ sys em.
Table 2summa izes he p o ocol s ack used a he p oposed a chi ec u e, wi h
IEEE802.11 as he s anda d o WiFi echnology and IEEE802.1Q as he s anda d o VLAN.
Table 2. Summa y o p o ocol s ack.
TCP/IP Laye s IoT P o ocols
Applica ion MQTT
T anspo TCP
In e ne IP 4
Da a Link IEEE802.11 (WiFi)
IEEE802.1Q (VLAN)
Message Model a he Edge/Fog Laye
Figu e 2summa izes he MQTT opics used o connec ing IEQ concen a o s wi h he
sma IEQ senso nodes. The IEQ concen a o holds he MQTT b oke o connec ing all
IEQ senso nodes in he a ea. No e ha published opics a e shown in o ange, whe eas sub-
sc ibed opics a e shown in blue. These opics a e associa ed wi h he
ollowing ope a ions:
•Con igu a ion o sma IEQ senso nodes
: The con igu a ion o one sma IEQ node
is ini ia ed a s a up ime, which eques s i s con igu a ion by means o a opic, con /ni,
whe e i ep esen s he node iden i ie numbe , ixed o e e y IEQ senso node. Upon
he ecep ion o his opic, he IEQ concen a o sends he a ailable con igu a ion o
ha speci ic node. Sma IEQ senso nodes ha e speci ic senso s a ached in di e en
layou s. So, he IEQ concen a o sends he con igu a ion o a speci ic node by means
o se e al published opics; he con ig/ni/nsens opic indica es he numbe o a ached
senso s, whe eas he speci ic con igu a ion o each senso is sen by he con ig/ni/sj
opic, which speci ies he ype o senso a ached o he pin j in he node i. These opics
a e ecei ed by he sma IEQ senso nodes, which sel -con igu e acco dingly.
•Rese o one/all IEQ senso nodes
: Occasionally, se e al issues may induce ope a -
ing p oblems a he IEQ moni o ing sys em. Fo his eason, he au ho s designed
a p ocedu e
o ese ing one o all a ached senso s. This ope a ion is ini ia ed by he
IEQ concen a o when i de ec s one o hese p oblems. The ese /ni opic, published
by he concen a o , is ecei ed by one speci ic sma IEQ node, igge ing he ese
p ocedu e. The ese /all opic was also included o simul aneously ese all connec ed
sma IEQ nodes.
Ene gies 2022,15, 2270 16 o 23
in Figu e 9c) ell du ing some ins ances o s a opening he doo and windows o c ea e
ai cu en s.
Ene gies 2022, 15, x FOR PEER REVIEW 17 o 25
Figu e 8. Dis ibu ion o he sma senso s a he Facul y o Enginee ing o Vi o ia-Gas eiz
(UPV/EHU).
Fo alida ion pu poses, he p o o ype o he IEQ moni o ing sys em was es ed in
di e en scena ios by aking di e se IEQ pa ame e s wi h di e en senso s. All es s we e
execu ed wi h a sampling pe iod o 5 min. The selec ed scena ios we e:
1. In a s a o ice
2. In a labo a o y du ing se e al labo a o y classes
3. In a class oom du ing an exam
4.1. In a S a O ice
This es shows how his sys em may be used o moni o ing he IEQ condi ions in a
s a o ice. Fou di e en sma IEQ senso nodes we e dis ibu ed in he o ice in di e -
en loca ions, moni o ing empe a u e (LM74), ela i e humidi y (SHT85), eCO2, and
TVOC (CCS811). This es was ca ied ou du ing almos 8 h, as shown in Figu e 9. IEQ
moni o ing da a we e acqui ed in he mon h o June o 2021.
Sma IEQ senso nodes p oduce di e en alues depending on loca ion. Some o
hem we e close o he windows, doo , elec onic de ices, o human beings. The cap u ed
esul s show ha he IEQ pa ame e s we e in adequa e anges. The alues cap u ed o
he CO2 concen a ion show ha i s alue is di e en depending on he loca ion (see Fig-
u e 9c). Fo example, Node 4 ( ed in Figu e 9c), was loca ed close o he window, whe eas
he o he nodes we e close o he s a . I can be app ecia ed ha he alues o Node 1
(blue in Figu e 9c) ell du ing some ins ances o s a opening he doo and windows o
c ea e ai cu en s.
Figu e 8.
Dis ibu ion o he sma senso s a he Facul y o Enginee ing o Vi o ia-Gas eiz
(UPV/EHU).
4.2. In a Labo a o y du ing Se e al Labo a o y Classes
This es was ca ied ou in a labo a o y in May 2021. This labo a o y is used o
eaching pu poses. The labo a o y was occupied wi h h ee consecu i e labo a o y sessions
o wo hou s, om 9:00 o 11:00; om 11:00 o 13:00; and om 13:00 o 15:00. The windows
we e always open. This es was aimed a measu ing empe a u e and humidi y pa ame e s
in he labo a o y o e alua ing he he mal com o o he s uden s.
Fi e sma IEQ senso nodes we e loca ed a di e en places. All i e IEQ nodes
acqui ed empe a u e (TMP37) bu only one o hem could also measu e ela i e humidi y
(SHT85). Node 4 ( ed in Figu e 10a) was nex o he window, whe eas he o he s we e
dis ibu ed along he labo a o y.
Ene gies 2022, 15, x FOR PEER REVIEW 18 o 25
(a) (b)
(c) (d)
Figu e 9. (a) Tempe a u e measu emen s (LM74 senso s); (b) Rela i e humidi y measu emen s
(SHT85); (c) eCO2 measu emen s (CCS811); (d) TVOC measu emen s (CCS811).
4.2. In a Labo a o y du ing Se e al Labo a o y Classes
This es was ca ied ou in a labo a o y in May 2021. This labo a o y is used o
eaching pu poses. The labo a o y was occupied wi h h ee consecu i e labo a o y ses-
sions o wo hou s, om 9:00 o 11:00; om 11:00 o 13:00; and om 13:00 o 15:00. The
windows we e always open. This es was aimed a measu ing empe a u e and humidi y
pa ame e s in he labo a o y o e alua ing he he mal com o o he s uden s.
Fi e sma IEQ senso nodes we e loca ed a di e en places. All i e IEQ nodes ac-
qui ed empe a u e (TMP37) bu only one o hem could also measu e ela i e humidi y
(SHT85). Node 4 ( ed in Figu e 10a) was nex o he window, whe eas he o he s we e
dis ibu ed along he labo a o y.
The ob ained esul s show ha he empe a u e condi ions in he labo a o y changed
conside ably, depending on whe e he nodes we e loca ed. In he mo ning, a 9:00, he
o e all empe a u e was be ween 25 and 28 deg ees, bu i eached 16 deg ees nex o he
window. Also, i can be app ecia ed ha Node 4 ( ed in Figu e 10a) p o ides he closes
alue o he ex e nal empe a u e, which inc eased du ing he day.
Figu e 9. Con .
Ene gies 2022,15, 2270 17 o 23
Ene gies 2022, 15, x FOR PEER REVIEW 18 o 25
(a) (b)
(c) (d)
Figu e 9. (a) Tempe a u e measu emen s (LM74 senso s); (b) Rela i e humidi y measu emen s
(SHT85); (c) eCO2 measu emen s (CCS811); (d) TVOC measu emen s (CCS811).
4.2. In a Labo a o y du ing Se e al Labo a o y Classes
This es was ca ied ou in a labo a o y in May 2021. This labo a o y is used o
eaching pu poses. The labo a o y was occupied wi h h ee consecu i e labo a o y ses-
sions o wo hou s, om 9:00 o 11:00; om 11:00 o 13:00; and om 13:00 o 15:00. The
windows we e always open. This es was aimed a measu ing empe a u e and humidi y
pa ame e s in he labo a o y o e alua ing he he mal com o o he s uden s.
Fi e sma IEQ senso nodes we e loca ed a di e en places. All i e IEQ nodes ac-
qui ed empe a u e (TMP37) bu only one o hem could also measu e ela i e humidi y
(SHT85). Node 4 ( ed in Figu e 10a) was nex o he window, whe eas he o he s we e
dis ibu ed along he labo a o y.
The ob ained esul s show ha he empe a u e condi ions in he labo a o y changed
conside ably, depending on whe e he nodes we e loca ed. In he mo ning, a 9:00, he
o e all empe a u e was be ween 25 and 28 deg ees, bu i eached 16 deg ees nex o he
window. Also, i can be app ecia ed ha Node 4 ( ed in Figu e 10a) p o ides he closes
alue o he ex e nal empe a u e, which inc eased du ing he day.
Figu e 9.
(
a
) Tempe a u e measu emen s (LM74 senso s); (
b
) Rela i e humidi y measu emen s
(SHT85); (c) eCO2measu emen s (CCS811); (d) TVOC measu emen s (CCS811).
Ene gies 2022, 15, x FOR PEER REVIEW 19 o 25
The IEQ moni o ing sys em also acqui ed he alues o he ela i e humidi y (Figu e
10b), which was always in he ange om 33% o 41 % and is conside ed accep able (bu
sligh ly d y) inside buildings.
Finally, i can be app ecia ed ha he e a e some missing alues in bo h igu es. This
happened a app oxima ely 11:15 and 13:00. The cause o his issue was ha he WiFi
communica ion was los o some ime a hese poin s. Howe e , he sys em was able o
econnec au oma ically.
(a)
(b)
Figu e 10. (a) Tempe a u e measu emen s (TMP37); (b) Rela i e humidi y measu emen s (SHT85).
Figu e 10. (a) Tempe a u e measu emen s (TMP37); (b) Rela i e humidi y measu emen s (SHT85).
Ene gies 2022,15, 2270 18 o 23
The ob ained esul s show ha he empe a u e condi ions in he labo a o y changed
conside ably, depending on whe e he nodes we e loca ed. In he mo ning, a 9:00, he
o e all empe a u e was be ween 25 and 28 deg ees, bu i eached 16 deg ees nex o he
window. Also, i can be app ecia ed ha Node 4 ( ed in Figu e 10a) p o ides he closes
alue o he ex e nal empe a u e, which inc eased du ing he day.
The IEQ moni o ing sys em also acqui ed he alues o he ela i e humidi y
(Figu e 10b)
,
which was always in he ange om 33% o 41 % and is conside ed accep able (bu sligh ly
d y) inside buildings.
Finally, i can be app ecia ed ha he e a e some missing alues in bo h igu es. This
happened a app oxima ely 11:15 and 13:00. The cause o his issue was ha he WiFi
communica ion was los o some ime a hese poin s. Howe e , he sys em was able o
econnec au oma ically.
4.3. In a Class oom du ing an Exam
This es was ca ied ou du ing an exam aken in he a e noon on 31 May 2021. The
exam ook wo and a hal hou s, om 16:00 o 18:30. In his case, he IEQ moni o ing sys em
acqui ed eCO
2
and empe a u e alues, using he CSS811 and TMP37 senso s, espec i ely.
The esul s o his es a e shown in Figu e 11. In his es , ou sma IEQ senso nodes
we e loca ed a di e en posi ions in he exam class oom. The acqui ed empe a u e alues
can be conside ed adequa e, since hey a e always o e 24 and below 28 deg ees.
Rega ding he alues o he eCO
2
, hey a e con inuously inc easing, al hough he
windows we e open in he exam. Calib a ion e o s could occu since he CSS811 senso has
no p o en o be e y p ecise. Rega dless, he acqui ed IEQ alues ecommend in oducing
addi ional ac ions in his scena io, such as in oducing auxilia y en ila ion sys ems o
imp o ing he quali y o he ai .
4.4. Discussion
The alues o di e en pa ame e s acqui ed om di e en sma IEQ sma nodes
a e co ec ly ecei ed and plo ed wi h he HMI applica ion. Only some imes was he
connec ion los due o a low WiFi signal bu i was apidly econnec ed wi h almos null
da a loss. I is also e lec ed ha depending on he posi ion o he node, he alues a e
di e en , due o he e ec o windows, doo s, elec onic de ices, o he p esence o humans.
Ene gies 2022, 15, x FOR PEER REVIEW 20 o 25
4.3. In a Class oom du ing an Exam
This es was ca ied ou du ing an exam aken in he a e noon on 31 May 2021. The
exam ook wo and a hal hou s, om 16:00 o 18:30. In his case, he IEQ moni o ing sys-
em acqui ed eCO2 and empe a u e alues, using he CSS811 and TMP37 senso s, espec-
i ely. The esul s o his es a e shown in Figu e 11. In his es , ou sma IEQ senso
nodes we e loca ed a di e en posi ions in he exam class oom. The acqui ed empe a-
u e alues can be conside ed adequa e, since hey a e always o e 24 and below 28 de-
g ees.
Rega ding he alues o he eCO2, hey a e con inuously inc easing, al hough he
windows we e open in he exam. Calib a ion e o s could occu since he CSS811 senso
has no p o en o be e y p ecise. Rega dless, he acqui ed IEQ alues ecommend in o-
ducing addi ional ac ions in his scena io, such as in oducing auxilia y en ila ion sys-
ems o imp o ing he quali y o he ai .
(a)
(b)
Figu e 11. (a) eCO2 measu emen s (CCS811); (b) Tempe a u e measu emen s (TMP37).
Figu e 11. Con .
Ene gies 2022,15, 2270 19 o 23
Ene gies 2022, 15, x FOR PEER REVIEW 20 o 25
4.3. In a Class oom du ing an Exam
This es was ca ied ou du ing an exam aken in he a e noon on 31 May 2021. The
exam ook wo and a hal hou s, om 16:00 o 18:30. In his case, he IEQ moni o ing sys-
em acqui ed eCO2 and empe a u e alues, using he CSS811 and TMP37 senso s, espec-
i ely. The esul s o his es a e shown in Figu e 11. In his es , ou sma IEQ senso
nodes we e loca ed a di e en posi ions in he exam class oom. The acqui ed empe a-
u e alues can be conside ed adequa e, since hey a e always o e 24 and below 28 de-
g ees.
Rega ding he alues o he eCO2, hey a e con inuously inc easing, al hough he
windows we e open in he exam. Calib a ion e o s could occu since he CSS811 senso
has no p o en o be e y p ecise. Rega dless, he acqui ed IEQ alues ecommend in o-
ducing addi ional ac ions in his scena io, such as in oducing auxilia y en ila ion sys-
ems o imp o ing he quali y o he ai .
(a)
(b)
Figu e 11. (a) eCO2 measu emen s (CCS811); (b) Tempe a u e measu emen s (TMP37).
Figu e 11. (a) eCO2measu emen s (CCS811); (b) Tempe a u e measu emen s (TMP37).
The designed IEQ moni o ing sys em wo ks p ope ly and can be used o ob ain da a
o de ec ing uncom o able si ua ions as well as p omo ing ac ing o educe ene gy con-
sump ion and imp o ing he heal h condi ions o s uden s and s a . These measu emen s
we e aken in he mon hs o May and June. In his case, he eason o he high empe a u es
was he ou side empe a u e and no he hea ing sys em. The acqui ed alues we e inside
he accep ed ange o alues mos o he ime. Howe e , al hough he au ho s conside ha
he p esen ed app oach is adequa e, i could be con enien o eplace he CSS811 senso
wi h one ha is mo e p ecise (and mo e expensi e) o measu ing he CO2concen a ion.
In summe , he ac ions ha could be aken o imp o ing he alues could be open-
ing/closing windows, u ning o emission sou ces, o educing he amoun o people
inside a oom. I necessa y, adding an ai condi ioning sys em o /and ai pu i ie could be
conside ed. In win e , he hea ing sys em should be adequa ely managed.
5. Conclusions and Fu u e Wo k
This pape in oduces a scalable IoT a chi ec u e based on he edge– og–cloud
pa adigm o moni o ing he Indoo s En i onmen al Quali y (IEQ) pa ame e s in public
and p i a e buildings. These sys ems allow o (1) ensu ing he he mal com o o he occu-
pan s, also imp o ing hei p oduc i i y, (2) in oducing ac ions o educe ene gy consump-
ion, con ibu ing o achie ing Sus ainable De elopmen Goals (SDG), and
(3) gua an eeing
ai quali y, which is a key conce n, as he COVID-19 pandemic has shown. These p oblems
occu equen ly in public o p i a e buildings as well as in indus ial acili ies.
The p oposed a chi ec u e implemen s he edge and og laye s by means o di e -
en de ices: sma IEQ senso nodes a he edge laye , which acqui e IEQ pa ame e s
locally, and IEQ concen a o s a he og laye , which collec he da a om se e al sma
IEQ senso nodes in one speci ic a ea and connec wi h cloud se ices. Edge– og–cloud
a chi ec u es ha e p o en o be adequa e o dealing wi h la ge amoun s o dis ibu ed
nodes in
Indus y 4.0
applica ions. This wo k p esen s i s applica ion o moni o ing IEQ
condi ions in buildings o g oups o buildings, such as hose on a uni e si y campus. The
hie a chical s uc u e o he a chi ec u e, as well as he use o TCP/IP echnologies, allows
o scaling he sys em o each di e en amoun s o nodes. Thus, IEQ Concen a o s a e
esponsible o collec ing he da a om speci ic a eas, p oducing a sys em ha may ope a e
au onomously. In addi ion, IEQ pa ame e s e ol e ela i ely slowly, so sampling imes in
he ange o minu es a e adequa e. Thus, since he ne wo k a ic o he sys em is ela i ely
low and non- ime-c i ical, a la ge numbe o de ices can be in oduced. The p oposed
sys em uses coope a i e ne wo k in as uc u es ha a e al eady deployed o ansmi ing
he moni o ed da a, which is a cos -e ec i e app oach.
Ene gies 2022,15, 2270 20 o 23
This sys em uses open-sou ce ha dwa e and so wa e echnologies, low-cos senso s,
and selec ed popula communica ion echnologies (WiFi, VLAN, TCP/IP and MQTT) o e
al eady deployed communica ion in as uc u es. The combina ion o all hese ea u es
p oduces a low-cos moni o ing sys em ha may be easily adap ed o measu e selec ed
IEQ pa ame e s. The p oposed a chi ec u e allows se e al le els o connec i i y. Single
boa d compu e s, such as Raspbe y Pi 3B+, we e used as IEQ concen a o s, whe eas
mic ocon olle boa ds, namely A duino MKR WiFi 1010 boa ds, we e used o he sma
IEQ senso nodes. Also, a PCB boa d was speci ically designed including di e en low-
cos senso s o measu ing he IEQ pa ame e s. Di e se s a e-o - he-a communica ion
echnologies o IoT applica ions we e used, namely WiFi, VLAN, In e ne , and MQTT. This
a chi ec u e could be used o moni o ing o he sys ems wi h sampling pe iods in he ange
o minu es, by changing he senso s and ixing he con igu a ion o he
acquisi ion sys em.
The p oposed IEQ moni o ing sys em is aimed a de ec ing uncom o able si ua ions
in buildings. Also, i may help o en o ce he SDGs by add essing u u e ac ions o imp o e
he e iciency o he hea ing and ai condi ioning sys ems. Inadequa e en i onmen al
condi ions may cause a educ ion in he p oduc i i y o building occupan s and may
e en lead o se e al diseases as occupan s spend long pe iods indoo s. Besides, wi h he
ad en o he COVID-19 pandemic, moni o ing he quali y o he ai indoo s has become
an impo an issue as he SARS-CoV-2 i us is sp ead h ough he ai . In his scena io,
moni o ing di e se a iables such as empe a u e, ela i e humidi y, eCO
2
, and TVOC may
help o keep hese issues unde con ol.
A p o o ype o he p oposed sys em was deployed a he Facul y o Enginee ing o
Vi o ia-Gas eiz (UPV/EHU). The p oposed IEQ moni o ing sys em is based on low-cos
ha dwa e and so wa e componen s. I was designed o be lexible and highly con igu able,
in o de o be deployed a buildings o di e en ypes and sizes wi h a ied cha ac e is ics.
This implemen a ion u ned ou o be success ul and cos -e ec i e.
The IEQ moni o ing sys em has been success ully es ed, p o ing ha i is able o
collec measu emen s om nodes dis ibu ed a ound di e en poin s o he acul y. Fo
es ing pu poses, se e al academic scena ios and con igu a ions we e chosen, achie ing
posi i e esul s. Namely, IEQ pa ame e s we e moni o ed in a s a o ice, in a eaching
labo a o y du ing se e al labo a o y classes, and in a class oom du ing an exam. The
sys em showed ha he wo king condi ions o he s a we e adequa e and ha some imes
he ela i e humidi y a he labo a o ies could be sligh ly low, and iden i ied some si ua ions
ha may equi e be e en ila ion.
Cu en ly, he au ho s a e wo king on de ec ing possible ac ions ha may imp o e
he IEQ condi ions o he occupan s a he acul y. Fo his pu pose, hey a e de eloping
au oma ic sys ems ha may s a /s op when needed, such as IoT au oma ic al es and
auxilia y en ila ion sys ems, o imp o ing bo h he mal condi ions and he quali y o
he ai . Also, he au ho s ha e de ec ed ha some senso s do no ope a e e y p ecisely
some imes, speci ically he CSS811 senso , so hey a e looking o ind a eplacemen . Finally,
he au ho s a e cu en ly de eloping ad anced cloud se ices aimed a analyzing la ge
se ies o da a.
Au ho Con ibu ions:
Concep ualiza ion, I.C. and E.A.; me hodology, I.C., J.M.G.-G. and O.B.; so -
wa e, A.E. and I.C.; alida ion, A.E. and I.C.; o mal analysis, I.C., J.M.G.-G. and P.F.B.; in es iga ion,
A.E., P.F.B. and I.C.; w i ing—o iginal d a p epa a ion, A.E. and I.C.; w i ing— e iew and edi ing,
I.C. and J.M.G.-G.; isualiza ion, A.E. and I.C.; supe ision, I.C.; p ojec adminis a ion, I.C., O.B. and
E.A.; All au ho s ha e ead and ag eed o he published e sion o he manusc ip .
Funding:
The au ho s wish o exp ess hei g a i ude, o suppo ing his wo k, o he Fundación
Vi al h ough p ojec VITAL21/05 and he Uni e si y o he Basque Coun y (UPV/EHU), h ough
he Campus Bizia Lab (CBL) p og am. Pa ial suppo has been also ecei ed om he Basque
Go e nmen , h ough p ojec EKOHEGAZ (ELKARTEK KK-2021/00092), he Dipu ación Fo al de
Ála a (DFA) h ough he p ojec CONAVANTER, and he UPV/EHU h ough he GIU20/063 g an .
Ins i u ional Re iew Boa d S a emen : No applicable.
Ene gies 2022,15, 2270 21 o 23
In o med Consen S a emen : No applicable.
Da a A ailabili y S a emen : No applicable.
Acknowledgmen s:
The au ho s wish o exp ess hei g a i ude o he Uni e si y o he Basque
Coun y (UPV/EHU), o suppo ing his wo k h ough he Campus Bizia Lab (CBL) p og am
and he Basque Go e nmen , h ough he p ojec EKOHEGAZ (ELKARTEK KK-2021/00092), he
Dipu ación Fo al de Ála a (DFA) h ough he p ojec CONAVANTER, and o he UPV/EHU h ough
he p ojec GIU20/063. Also, hey exp ess g a i ude o Fundación VITAL o suppo ing his wo k
wi h VITAL21/05 p ojec .
Con lic s o In e es : The au ho s decla e no con lic o in e es .
Re e ences
1.
Chen, B.; Wan, J.; Shu, L.; Li, P.; Mukhe jee, M.; Yin, B. Sma Fac o y o Indus y 4.0: Key Technologies, Applica ion Case, and
Challenges. IEEE Access 2017,6, 6505–6519. [C ossRe ]
2.
Candell, R.; Kashe , M.; Liu, Y.; Lee, K.B.; Fou ou, S. Indus ial wi eless sys ems guidelines: P ac ical conside a ions and
deploymen li e cycle. IEEE Ind. Elec on. Mag. 2018,12, 6–17. [C ossRe ]
3.
Mazon Oli o, B.; Pan, A. In e ne o Things: S a e-o - he-a , Compu ing Pa adigms and Re e ence A chi ec u es. IEEE La . Am.
T ans. 2021,20, 49–63. [C ossRe ]
4.
Pa kinson, T.; Pa kinson, A.; de Dea , R. Con inuous IEQ moni o ing sys em: Con ex and de elopmen . Build. En i on.
2019
,
149, 15–25. [C ossRe ]
5.
Cal o, I.; Gil-Ga cía, J.M.; Recio, I.; López, A.; Quesada, J. Building IoT Applica ions wi h Raspbe y Pi and Low Powe IQRF
Communica ion Modules. Elec onics 2016,5, 54. [C ossRe ]
6.
Dalampi a, E.-S.; Nas is, S.A. Mapping Sus ainable De elopmen Goals: A ne wo k analysis amewo k. Sus ain. De .
2020
,
28, 46–55. [C ossRe ]
7.
Yang, L.; Yan, H.; Lam, J.C. The mal com o and building ene gy consump ion implica ions—A e iew. Appl. Ene gy
2014
,
115, 164–173. [C ossRe ]
8.
Cal o, I.; Gil-Ga cía, J.M.; Apiñaniz, E.; Escude o, C.; Ga cía-Ade a, A.J.; Mesanza, A.; Gas ón, M. Design o a PBL expe ience in
he ield o sus ainabili y o indus ial in o ma ics. Ad . In ell. Sys . Compu . 2021,1266, 338–347. [C ossRe ]
9.
And eo i, M.; Calzola i, M.; Da oli, P.; Pe ei a, L.D.; Lucchi, E.; Malagu i, R. Design and cons uc ion o a new me e ing ho box
o he in si u hyg o he mal measu emen in dynamic condi ions o his o ic mason ies. Ene gies 2020,13, 2950. [C ossRe ]
10.
Coulby, G.; Clea , A.; Jones, O.; God ey, A. A scoping e iew o echnological app oaches o en i onmen al moni o ing. In . J.
En i on. Res. Public Heal h 2020,17, 3995. [C ossRe ]
11.
Pie og ande, M.C.; Casa i, L.; Dema ia, G.; Russo, M. Indoo ai quali y in domes ic en i onmen s du ing pe iods close o i alian
COVID-19 lockdown. In . J. En i on. Res. Public Heal h 2021,18, 4060. [C ossRe ] [PubMed]
12.
Bu idge, H.C.; Bhaga , R.K.; S e le , M.E.J.; Kuma , P.; De Mel, I.; Demis, P.; Ha , A.; Johnson-Llambias, Y.; King, M.F.;
Klymenko, O.; e al.
The en ila ion o buildings and o he mi iga ing measu es o COVID-19: A ocus on win e ime. P oc. R.
Soc. A 2021,477, 20200855. [C ossRe ] [PubMed]
13.
Fi ouzi, F.; Fa ahani, B.; Ma inšek, A. The con e gence and in e play o edge, og, and cloud in he AI-d i en In e ne o Things
(IoT). In . Sys . 2021, in p ess. [C ossRe ]
14.
Anda gie, M.S.; Touchie, M.; O’B ien, W. A e iew o ac o s a ec ing occupan com o in mul i-uni esiden ial buildings. Build.
En i on. 2019,160, 106182. [C ossRe ]
15.
Co gna i, S.P.; Filippi, M.; Viazzo, S. Pe cep ion o he he mal en i onmen in high school and uni e si y class ooms: Subjec i e
p e e ences and he mal com o . Build. En i on. 2007,42, 951–959. [C ossRe ]
16.
Saini, J.; Du a, M.; Ma ques, G. Indoo ai quali y moni o ing sys ems based on in e ne o hings: A sys ema ic e iew. In . J.
En i on. Res. Public Heal h 2020,17, 4942. [C ossRe ] [PubMed]
17.
ANSI/ASHRAE S anda d 55-2010; The mal En i onmen al Condi ions o Human Occupacy. Ame ican Socie y o Hea ing,
Re ige a ing and Ai -Condi ioning Enginee s, Inc.: A lan a, GA, USA, 2011. A ailable online: h p://a co-h ac.i /wp-con en /
uploads/2015/11/ASHRAE-55-2010.pd (accessed on 7 Ma ch 2022).
18.
RESET Ai S anda d o Comme cial In e io s 2.0, RESETTM. A ailable online: h ps://www. ese .build/s anda d/ai (accessed
on 7 Ma ch 2022).
19.
ISO 7730:2005 E gonomics o he The mal En i onmen —Analy ical De e mina ion and In e p e a ion o The mal Com o Using
Calcula ion o he PMV and PPD Indices and Local The mal Com o C i e ia. A ailable online: h ps://www.iso.o g/s anda d/
39155.h ml (accessed on 7 Ma ch 2022).
20.
Yi, W.; Lo, K.; Mak, T.; Leung, K.; Leung, Y.; Meng, M. A Su ey o Wi eless Senso Ne wo k Based Ai Pollu ion Moni o ing
Sys ems. Senso s 2015,15, 31392–31427. [C ossRe ] [PubMed]
21.
Pe ez, A.O.; Bie e , B.; Scholz, L.; Wöllens ein, J.; Palze , S. A wi eless gas senso ne wo k o moni o indoo en i onmen al
quali y in schools. Senso s 2018,18, 4345. [C ossRe ]
Ene gies 2022,15, 2270 22 o 23
22.
Dang, C.T.; Seide e , A.; And é, E. Theodo : A s ep owa ds sma home applica ions wi h elec onic noses. In P oceedings o
he ACM In e na ional Con e ence P oceeding Se ies, 5 h in e na ional Wo kshop on Senso based Ac i i y Recogni ion and
In e ac ion, iWOAR 2018, Be lin, Ge many, 20–21 Sep embe 2018. [C ossRe ]
23.
Gomes, J.B.A.; Rod igues, J.J.P.C.; Rabêlo, R.A.L.; Kuma , N.; Kozlo , S. IoT-enabled gas senso s: Technologies, applica ions, and
oppo uni ies. J. Sens. Ac ua o Ne w. 2019,8, 57. [C ossRe ]
24.
Seide e , A.; Aslan, I.; Dang, C.T.; And é, E. Indoo ai quali y and wellbeing—Enabling awa eness and sensi i i y wi h
ambien IOT displays. In P oceedings o he 15 h Eu opean Con e ence on Ambien In elligence, AmI 2019, Rome, I aly,
13–15 No embe 2019; Volume 11912, pp. 266–282. [C ossRe ]
25.
Yang, C.-T.; Chen, S.-T.; Den, W.; Wang, Y.-T.; K is iani, E. Implemen a ion o an In elligen Indoo En i onmen al Moni o ing and
managemen sys em in cloud. Fu u e Gene . Compu . Sys . 2019,96, 731–749. [C ossRe ]
26.
Zhang, X.; Zhao, Y.; Zhao, W.; Xu, W.; Ji, W. A wi eless senso ne wo ks-based in elligen sys em o lib a y ai quali y moni o ing.
In . J. Online Eng. 2016,12, 76–79. [C ossRe ]
27.
Youness, A.; Mus apha, H.; Sul an, A.A.; Ahmed, A.S. Low-cos da a acquisi ion s a ion o mic oclima e de ec ion and public
heal h isks managemen : A case s udy o Casablanca. In . J. Online Biomed. Eng. 2020,16, 31–40. [C ossRe ]
28.
Schia on, M.; Adami, L.; Maga il, E.; Ragazzi, M. Indoo CO
2
: Po en ial c i icali ies and solu ions. In P oceedings o he MATEC
Web o Con e ences, 9 h In e na ional Con e ence on Manu ac u ing Science and Educa ion: T ends in New Indus ial Re olu ion,
Sibiu, Romania, 5–7 June 2019; Volume 290, p. 12026. [C ossRe ]
29.
He, Y.; Tan, M. Implemen a ion o ai -condi ioning con ol echnology based on wi eless senso . In . J. Online Eng.
2017
,
13, 91–99. [C ossRe ]
30.
Ba mpa esos, N.; Papadaki, D.; Ka alis, M.; Famelia i, K.; Assimakopoulos, M.N. In si u measu emen s o ene gy consump ion
and indoo en i onmen al quali y o a p e- e o i ed s uden do mi o y in A hens. Ene gies 2019,12, 2210. [C ossRe ]
31.
Ma ín-Ga ín, A.; Millán-Ga cía, J.A.; Baï i, A.; Millán-Medel, J.; Sala-Liza aga, J.M. En i onmen al moni o ing sys em based on
an Open Sou ce Pla o m and he In e ne o Things o a building ene gy e o i . Au om. Cons . 2018,87, 201–214. [C ossRe ]
32.
Ka ami, M.; McMo ow, G.V.; Wang, L. Con inuous moni o ing o indoo en i onmen al quali y using an A duino-based da a
acquisi ion sys em. J. Build. Eng. 2018,19, 412–419. [C ossRe ]
33.
Salamone, F.; Belussi, L.; Danza, L.; Galanos, T.; Ghelle e, M.; Me oni, I. Design and de elopmen o a nea ablewi eless sys em o
con ol indoo ai quali y and indoo ligh ing quali y. Senso s 2017,17, 1021. [C ossRe ]
34.
Menacho, A.; Cas o, M.; Pé ez, C. A duino-based wa e analysis pocke lab. In P oceedings o he 2021 Wo ld Enginee ing
Educa ion Fo um/Global Enginee ing Deans Council, WEEF/GEDC 2021, Mad id, España, 15–18 No embe 2021. [C ossRe ]
35.
Ali, A.S.; Co é, C.; Heida inejad, M.; S ephens, B. Elemen al: An Open-Sou ce Wi eless Ha dwa e and So wa e Pla o m o
Building Ene gy and Indoo En i onmen al Moni o ing and Con ol. Senso s 2019,19, 4017. [C ossRe ]
36.
Al-Fuqaha, A.; Guizani, M.; Mohammadi, M.; Aledha i, M.; Ayyash, M. In e ne o Things: A Su ey on Enabling Technologies,
P o ocols, and Applica ions. IEEE Commun. Su . Tu o . 2015,17, 2347–2376. [C ossRe ]
37.
Paniagua, E.; Macazana, J.; Lopez, J.; Ta illo, J. IoT-based empe a u e moni o ing o buildings he mal com o analysis. In
P oceedings o he 2019 IEEE 26 h In e na ional Con e ence on Elec onics, Elec ical Enginee ing and Compu ing, INTERCON
2019, Lima, Pe u, 12–14 Augus 2019. [C ossRe ]
38.
Va gas-Salgado, C.; Aguila-Leon, J.; Chiñas-Palacios, C.; Hu ado-Pe ez, E. Low-cos web-based Supe iso y Con ol and Da a
Acquisi ion sys em o a mic og id es bed: A case s udy in design and implemen a ion o academic and esea ch applica ions.
Heliyon 2019,5, e02474. [C ossRe ]
39.
Po alo, J.M.; González, I.; Calde ón, A.J. Moni o ing sys em o acking a p gene a o in an expe imen al sma mic og id:
An open-sou ce solu ion. Sus ainabili y 2021,13, 8182. [C ossRe ]
40.
De Melo, G.C.G.; To es, I.C.; de A aújo, Í.B.Q.; B i o, D.B.; Ba boza, E.A. A low-cos io sys em o eal- ime moni o ing o
clima ic a iables and pho o ol aic gene a ion o sma g id applica ion. Senso s 2021,21, 3293. [C ossRe ] [PubMed]
41.
Lucchi, E.; Pe ei a, L.D.; And eo i, M.; Malagu i, R.; Cennamo, D.; Calzola i, M.; F ighi, V. De elopmen o a compa ible, low
cos and high accu a e conse a ion emo e sensing echnology o he hyg o he mal assessmen o his o ic walls. Elec onics
2019,8, 643. [C ossRe ]
42.
Kanal, A.K.; Ko acshazy, T. IoT solu ion o assessing he indoo ai quali y o educa ional acili ies. In P oceedings o he 2019
20 h In e na ional Ca pa hian Con ol Con e ence ICCC 2019, K akow-Wieliczka, Poland, 26–29 May 2019. [C ossRe ]
43.
A oyo, P.; Meléndez, F.; Suá ez, J.I.; He e o, J.L.; Rod íguez, S.; Lozano, J. Elec onic nose wi h digi al gas senso s connec ed ia
blue oo h o a sma phone o ai quali y measu emen s. Senso s 2020,20, 786. [C ossRe ] [PubMed]
44.
Sakuma, Y.; Nishi, H. Es ima ion o building he mal pe o mance using simple senso s and ai condi ione s. Ene gies
2019
,
12, 2950. [C ossRe ]
45. Jin-Feng, L.; Shun, C. A low-cos wi eless wa e quali y au o-moni o ing sys em. In . J. Online Eng. 2015,11, 37–41. [C ossRe ]
46.
Baig, F.; Mahmood, A.; Ja aid, N.; Razzaq, S.; Khan, N.; Saleem, Z. Sma home ene gy managemen sys em o moni o ing and
scheduling o home appliances using zigbee. J. Basic Appl. Sci. Res. 2013,3, 880–891.
47.
F oiz-Míguez, I.; Fe nández-Ca amés, T.M.; F aga-Lamas, P.; Cas edo, L. Design, implemen a ion and p ac ical e alua ion o
an io
home au oma ion sys em o og compu ing applica ions based on MQTT and ZigBee-WiFi senso nodes. Senso s
2018
,
18, 2660. [C ossRe ]
Ene gies 2022,15, 2270 23 o 23
48.
Mois, G.; Sanisla , T.; Folea, S.C. A Cybe -Physical Sys em o En i onmen al Moni o ing. IEEE T ans. Ins um. Meas.
2016
,
65, 1463–1471. [C ossRe ]
49.
Azemi, S.N.; Loon, K.W.; Ami , A.; Kamal udin, M. An IoT-Based Ala m Ai Quali y Moni o ing Sys em. J. Phys. Con . Se .
2021
,
1755, 012035. [C ossRe ]
50.
Jamal, A.; Al Na ayanasamy, D.D.; Mohd Zaki, N.Q.; Abbas Helmi, R.A. La ge Hall Tempe a u e Moni o ing Po al. In
P oceedings o he 2019 IEEE In e na ional Con e ence on Au oma ic Con ol and In elligen Sys ems, I2CACIS 2019, Selango ,
Malaysia, 29 June 2019. [C ossRe ]
51.
Al-Mas i, E.; Kalyanam, K.R.; Ba s, J.; Kim, J.; Singh, S.; Vo, T.; Yan, C. In es iga ing Messaging P o ocols o he In e ne o
Things (IoT). IEEE Access 2020,8, 94880–94911. [C ossRe ]
52.
Naik, N. Choice o e ec i e messaging p o ocols o IoT sys ems: MQTT, CoAP, AMQP and HTTP. In P oceedings o he 2017
IEEE In e na ional Symposium on Sys ems Enginee ing, ISSE 2017, Vienna, Aus ia, 11–13 Oc obe 2017. [C ossRe ]
53. Jaika , S.P.; Iye , K.R. A Su ey o Messaging P o ocols o IoT Sys ems. In . J. Ad . Manag. Technol. Eng. Sci. 2018,8, 510–514.
54.
Alae jan, A.; Kim, D.K.; Ming, H.; Kim, H. Con igu able DDS as uni o m middlewa e o da a communica ion in sma g ids.
Ene gies 2020,13, 1839. [C ossRe ]
55.
Cal o, I.; Pé ez, F.; E xebe ia-Agi iano, I.; Ga cía de Albéniz, O. Designing High Pe o mance Fac o y Au oma ion Applica ions
on Top o DDS. In . J. Ad . Robo . Sys . 2013,10, 205. [C ossRe ]
56. Mish a, B.; Ke esz, A. The use o MQTT in M2M and IoT sys ems: A su ey. IEEE Access 2020,8, 201071–201086. [C ossRe ]
57.
Delgado, A.; Huamaní, E.L.; Aguila-Ruiz, B. Design o an IoT domo ic sys em using he MQTT p o ocol. In . J. Ad . T ends
Compu . Sci. Eng. 2020,9, 4811–4818. [C ossRe ]
58.
Ca alcan e, T.L.; Louzada, D.R.; da Sil a, A.; Mon ei o, E.C. Mul ipa ame ic measu ing sys em o a mosphe ic moni o ing.
J. Phys. Con . Se . 2021,1826, 012020. [C ossRe ]
59.
Vanus, J.; Go jani, O.M.; Bilik, P. No el p oposal o p edic ion o CO
2
cou se and occupancy ecogni ion in in elligen buildings
wi hin IoT. Ene gies 2019,12, 4541. [C ossRe ]
60.
Indus ial Wi eless Remo e Tempe a u e Moni o ing Sys ems|E-Con ol Sys ems. A ailable online: h ps://econ olsys ems.com/
(accessed on 7 Sep embe 2021).
61.
Labo a o y Tempe a u e Moni o ing Sys ems—Sensoscien i ic. A ailable online: h ps://sensoscien i ic.com/labo a o y-
empe a u e-moni o ing-sys ems/ (accessed on 7 Sep embe 2021).
62.
Remo e Tempe a u e & Humidi y Moni o ing|Con olByWeb. A ailable online: h ps://www.con olbyweb.com/ empe a u e/
(accessed on 7 Sep embe 2021).
63.
Kaliwo, A.; Pini olo, J.; Mikeka, C. Real-Time, Web-based Tempe a u e Moni o ing Sys em o Cold Chain Managemen in
Malawi. J. Wi el. Ne w. Commun. 2017,7, 53–58.
64.
Analog De ices, Low Vol age Tempe a u e Senso s, TMP35/TMP36/TMP37. A ailable online: h ps://www.analog.com/media/
en/ echnical-documen a ion/da a-shee s/TMP35_36_37.pd (accessed on 28 Feb ua y 2022).
65.
Texas Ins umen s, LM74 SPI/Mic owi e 12-Bi Plus Sign Tempe a u e Senso . A ailable online: h ps://www. i.com/p oduc /
LM74# ech-docs (accessed on 28 Feb ua y 2022).
66.
Sensi ion, Da ashee SHT85 Humidi y and Tempe a u e Senso . A ailable online: h ps://sensi ion.com/media/documen s/4B4
0CEF3/61642381/Sensi ion_Humidi y_Senso s_SHT85_Da ashee .pd (accessed on 28 Feb ua y 2022).
67.
AMS, CCS811 Ul a-Low Powe Digi al Gas Senso o moni o ing Indoo Ai Quali y. A ailable online: h ps://cdn.spa k un.
com/asse s/lea n_ u o ials/1/4/3/CCS811_Da ashee -DS000459.pd (accessed on 28 Feb ua y 2022).
