IT 18 018
Examensarbete 45 hp
Juni 2018
Sensor Network for Real-Time
Air Quality Measurement
Teknisk- naturvetenskaplig fakultet UTH-enheten Besöksadress: Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0 Postadress: Box 536 751 21 Uppsala Telefon: 018 – 471 30 03 Telefax: 018 – 471 30 00 Hemsida: http://www.teknat.uu.se/student
Abstract
Sensor Network for Real-Time Air Quality
Measurement
Rudolfs Eriks Agrens
In this research, a sensor network for air quality measurement is developed.
There is a need for air quality assessment in rapidly developing urban areas. Prolonged real-time monitoring can provide a direct feedback on how increasing traffic and construction work affect air quality. Measures can then be taken to mitigate pollution, reducing health hazards and maintaining the standard of air quality in all areas of the city. A low-cost sensor node is required to be able to cover a larger area using more sensor nodes.
Methods for measuring humidity, temperature, particle concentration, as well as the gasses carbon monoxide, nitrogen dioxide and ozone were investigated and a sensor node designed by choosing suitable sensors.
Field-testing of the sensor nodes was performed in the center of Uppsala city in collaboration with Uppsala Kommun and Uppsala Stadsteater. The commercially available Air Quality Egg was used as a reference method for sensor calibration and verification of measurement accuracy.
The results show that the readings match the reference method well. Even low gas concentrations can be measured with good accuracy.
There is future interest from Uppsala Kommun to expand the sensor network for measurements in other areas than the city center.
!
DECLARATION!
I!hereby!certify!that!this!report!constitutes!my!own!product,!that!where!the!language!of!others!is!set! forth,! quotation! marks! so! indicate,! and! that! appropriate! credit! is! given! where! I! have! used! the! language,! ideas,!expressions!or!writings!of!another.! I!declare!that!the!report!describes!original!work!that!has!not!previously!been!presented!for!the!award!of! any!other!degree!of!any!institution.! Both!the!abstract!and!the!content!of!this!report!are!intellectual!property!of!Rudolfs!Agrens,!and!other! parts!of!the!research!are!intellectual!property!of!Aji!John!Pushpam.!This!research!is!not!to!be!used!for!other! research!or!commercial!purposes!without!the!author’s!written!consent.! ! ! ! Signed,!! ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! RUDOLFS!ERIKS!AGRENS!
!
ACKNOWLEDGEMENTS! !
This! research! was! conducted! with! Aji! John! Pushpam,! whose! thesis! is! written! concerning! the! wireless! communication!of!the!network!and!data!trends.! I!would!like!to!thank!my!supervisors!Dr.!Edith!Ngai!and!Dr.!Christian!Rohner!for!their!support!and!technical! advice.!I!would!like!to!thank!Dr.!Roland!Bol!for!valuable!feedback!during!the!writing!of!the!thesis.!I!would! also!like!to!thank!Dr.!Uwe!Zimmerman!for!advice!when!designing!the!electronics!circuit!for!the!sensor!node.! The!research!would!not!have!been!possible!without!their!expertise!and!their!initiative!to!support!the!thesis.!! I!would!also!like!to!express!gratitude!to!the!staff!of!Uppsala!Kommun!and!to!the!staff!of!Uppsala!Stadsteater! for! their! assistance! in! installing! the! sensor! nodes! and! for! lending! their! facilities! for! performing! the! measurements.!
Lastly,!I!would!like!to!thank!my!family!for!all!their!support.! The!research!was!funded!by!Uppsala!University,!Sweden.! !
! TABLE!OF!CONTENTS! List!of!Tables!...!9! List!of!Figures!...!10! 1!!Rationale!...!11! 1.1!Introduction!...!11! 1.2!Problem!definition!...!11! 1.2.1!Other!technical!constraints!and!specifications!...!12! 1.3!Project!scope!and!research!question!...!12! 1.4!Project!stakeholders!...!13! 2!!Situational!analysis!and!related!work!...!14! 2.1!Situational!Analysis!...!14! 2.2!Background!on!air!quality!measurement!...!16! 2.3!Related!work!...!16! 3!–!Conceptual!model!...!18! 3.1!Microcontroller!...!18! 3.2!Temperature!and!humidity!sensor!...!18! 3.3!Particle!concentration!sensor!...!19! 3.4!Ozone!and!nitrogen!dioxide!sensors!...!19! 3.5!Carbon!monoxide!sensor!...!22! 3.6!Cost!estimation!...!23! 4!–!Research!design!...!24! 4.1!Measurement!location!and!setup!...!24!
5.1.1!Circuit!design!...!26! 5.1.2!Casing!and!installation!...!27! 5.2!Measurement!results!...!28! 5.2.1!Carbon!monoxide!sensor!readings!...!28! 5.2.2!Nitrogen!dioxide!sensor!readings!...!29! 5.2.3!Ozone!sensor!readings!...!34! 5.2.4!Power!consumption!...!37! 6!/!Conclusions!...!39! 7!–!Future!work!and!recommendations!...!40! References!...!41! APPENDIX!!...!45! ! ! ! !
List%of%Tables%
Table!1:!Newly!registered!vehicles!in!Uppsala!by!type,!2015!...!14!
Table!2:!Cost!estimation!and!parts!list!per!sensor!node!...!23!
Table!3:!Temperature!and!average!error,!NO2!readings,!Uppsala!Theater!node!...!32!
Table!4:!Power!consumption!of!sensor!node!in!active!mode!...!37!
Table!5:!Power!consumption!of!sensor!node!in!sleep!mode!...!37!
Table!6:!Battery!runtime!of!sensor!node!...!38!
Table!7:!Battery!runtime!of!sensor!using!sensor!board!disconnecting!...!38!
!List%of%Figures%
Figure!1:!Air!quality!measurement!locations!in!Europe![9]!...!15!
Figure!2:!Ozone!sensor!calibration!curve!...!21!
Figure!3:!Nitrogen!dioxide!calibration!curve!...!21!
Figure!4:!Circuit!schematic!of!the!power!saving!node!version!...!26!
Figure!5:!Sensor!node!installed!at!Uppsala!Kommun!...!27!
Figure!6:!Node!locations,!sensor!nodes!in!blue,!existing!unit!of!the!Kommun!in!yellow!...!27!
Figure!7:!CO!reading!comparison!between!node!at!theater!and!reference!Air!Quality!Egg!...!28!
Figure!8:!NO2!reading!comparison!between!node!at!theater!and!reference!Air!Quality!Egg!...!29!
Figure!9:!Hourly!averages!of!NO2!readings!at!Uppsala!Theater!...!30!
Figure!10:!NO2!error!vs!temperature,!Uppsala!Theater!node!...!30!
Figure!11:!NO2!error!vs!humidity,!Uppsala!Theater!node!...!31!
Figure!12:!Temperature!compensated!NO2!measurements,!Uppsala!Theater!node!...!31!
Figure!13:!NO2!concentration,!Kungsgatan!and!Vaksalagatan!nodes!...!32!
Figure!14:!Temperature!compensated!NO2!measurements,!Kungsgatan!and!Vaksalagatan!nodes!...!33!
Figure!15:!O3!readings,!Uppsala!Theater!node!...!34!
Figure!16:!Temperature!compensated!O3!readings,!Uppsala!Theater!node!...!34!
Figure!17:!O3!readings,!Kungsgatan!node!...!35!
Figure!18:!Temperature!compensated!O3!readings,!Kungsgatan!node!...!35!
%
1%%Rationale%
!1.1#Introduction#
There!exists!a!need!for!a!low/cost!air!quality!measurement!network!that!could!give!a!broader!overview!of!air! quality!in!Uppsala!city.!Uppsala!Kommun!have!expressed!an!interest!in!this!and!there!are!project!initiatives! for!building!such!a!network.! Currently!there!aren’t!any!existing!city/wide!scale!networks!to!monitor!air!quality!in!Sweden.!Measurements! are!however!done!at!local!points.!In!the!case!of!Uppsala,!there!are!two!units!in!the!city!measuring!particle! count.!However,!these!cannot!give!detailed!enough!information,!both!because!of!the!small!area!covered!and! because!only!a!single!parameter!(particle!count)!is!measured.!! For!a!better!understanding!of!air!quality!in!Uppsala!city,!a!larger!network!is!needed!with!more!parameters! measured.!The!sensor!nodes!should!be!reliable!and!low/cost.!The!cost!of!currently!available!commercial!air! quality!sensor!nodes!is!too!high!to!create!a!large!network.!Therefore!the!goal!of!this!project!is!to!design!a! sensor! node! to! measure! several! air! quality! related! parameters! reliably! while! being! a! low/cost! solution.! It! should!meet!the!cost!margin!and!have!low!maintenance!requirements!in!order!to!be!expanded!into!a!larger! sensor! network! in! the! future.! The! research! is! conducted! in! collaboration! with! Uppsala! University,! while! Uppsala! Kommun! and! Uppsala! Stadsteater! provide! locations! for! placing! sensor! nodes.! The! research! was! performed!jointly!with!Aji!John!Pushpam,!whose!focus!is!on!data!analysis!and!wireless!communication,!and! for!these!parts,!references!will!be!given!to!his!thesis.1.2#Problem#definition#
The! measurement! parameters! –! humidity,! temperature,! particle! count,! and! the! gasses! carbon! monoxide! (CO),! nitrogen! dioxide! (NO2)! and! ozone! (O3)! –! were! selected! based! on! information! by! the! United! States!
Environmental!Protection!Agency!(EPA)![10].!The!considered!criteria!for!the!sensor!node!were:! 1.!Reliable!sensors!(that!other!research!refers!to)! 2.!Relatively!low/cost!sensors! 3.!Weatherproof!casing! 4.!Compact!design!and!ease!of!installation! ! The!required!measurement!range!for!relative!humidity!is!at!least!20%!to!80%!and!for!temperature!it!is!at! least!from!/20!to!+40!degrees!Celsius.!It!is!assumed!that!weather!conditions!will!mostly!be!in!this!range!in! the!measurement!location!Uppsala,!Sweden.! For!the!other!sensors,!the!exact!performance!is!harder!to!quantify!and!will!have!to!be!investigated!in!field! measurements.!Unfortunately,!some!heated!gas!sensors!pose!and!additional!challenge!in!the!design,!as!it!is! possible! that! these! are! not! designed! for! low! temperatures.! For! conditions! in! winter,! some! sensors! may! therefore!become!unusable.!
1.2.1%Other%technical%constraints%and%specifications%
! There!are!other!industrial!solutions!for!air!quality!measurement!that!use!one!or!multiple!fans!to!exchange!air! between!the!inside!and!outside!of!the!casing.!This!option!was!considered,!but!was!decided!against!since!the! mechanical!design!of!the!sensor!node!is!a!large!research!topic!in!itself!that!would!not!be!possible!to!cover! due!to!the!limited!time!of!the!thesis.!The!simpler!option!was!chosen,!making!a!large!opening!in!the!casing!for! airflow.!The!measurement!results!will!of!course!be!evaluated!in!comparison!with!a!reference!method!/!Air! Quality!Egg![39]/!and!based!on!this,!the!necessity!of!a!fan!may!or!may!not!be!suggested!for!future!research.! The!sensor!nodes!will!use!mains!power!instead!of!battery!power.!Focusing!on!designing!a!battery/powered! node!would!greatly!limit!the!choice!of!sensors!and!therefore!possibly!also!performance.!Many!gas!sensors! use! heating! as! their! measurement! principle,! which! results! in! higher! power! consumption.! Sensors! will! be! chosen!primarily!based!on!their!measurement!characteristics!and!cost,!rather!than!power!usage.!The!power! consumption!will!however!be!evaluated!in!detail,!and!it!will!be!assessed!whether!a!battery!powered!node! would!be!a!possibility.!1.3#Project#scope#and#research#question#
! The!main!research!question,!based!on!the!aforementioned!requirements,!is:! !Can!a!reliable,!low-cost!sensor!node!be!designed!for!air!quality!measurement?!
! Low!cost!is!defined!as!2000!SEK!in!material!costs,!and!not!exceeding!3000!SEK.!This!price!range!was!chosen! because! it! is! lower! than! similar! handheld! gas! monitoring! solutions! and! this! would! make! it! affordable! to! increase! the! number! of! nodes! and! expand! into! a! larger! sensor! network.! Reliability! refers! to! the! measurement!error,!and!the!exact!error!margin!is!subject!to!this!research!and!will!be!evaluated.!As!a!goal!for! the!research,!the!measurement!will!be!considered!sufficiently!accurate!if!the!average!error!is!less!than!20! percent.! ! Sub/questions!based!on!the!problem!definition!are:! !Can!air!quality!sensor!calibration!be!carried!out!as!field-calibration?!
!The! method! of! the! research! is! to! place! the! nodes! in! the! desired! measurement! location! (Uppsala! centre),! gather!measurements,!and!calibrate!the!readings!afterwards!by!using!the!Air!Quality!Egg!as!reference.!Field/ calibration!will!be!carried!out,!as!opposed!to!calibrating!in!a!lab!environment.!In!field/calibration,!none!of!the! measured! variables! are! held! fixed! but! instead! are! allowed! to! vary! as! they! normally! would! in! the! (urban)! environment.!Calibration!will!be!performed!by!first!scaling!the!obtained!data!to!match!that!of!the!reference! Air! Quality! Egg! and! then! applying! calibration! curves! given! by! sensor! manufacturers.! Temperature! and! humidity! calibration! will! not! be! set! as! a! requirement! as! this! would! require! a! lab! environment! with! fixed! variables!and!exceeds!the!project!scope.!However,!the!influence!of!humidity!and!temperature!on!the!sensor! readings!will!be!investigated!in!detail.!The!exact!calibration!procedures!will!be!discussed!further!in!the!thesis.
Are!the!sensor!measurements!accurate!and!repeatable!enough!to!evaluate!air!quality?!
!
The! readings! will! be! considered! accurate! if! they! are! of! comparable! quality! with! the! Air! Quality! Egg.! The! correlation!between!the!readings!of!our!sensor!node!and!those!of!the!Air!Quality!Egg!will!be!investigated.! The!readings!will!be!considered!repeatable!if!during!a!measurement!period!of!24!hours!or!more!there!is!no! considerable!decrease!in!accuracy.!
Can!the!sensor!node!in!theory!be!designed!to!be!battery-operated?!
!
The! nodes! will! be! operated! directly! using! mains! power,! and! building! battery/operated! nodes! is! excluded! from!the!scope!of!the!research.!The!nodes!are!expected!to!consume!a!high!amount!of!power!due!to!several! of!the!sensors!using!heating!in!their!operational!principle.!However,!power!consumption!will!be!researched! in! detail.! A! battery/operated! node! will! be! considered! a! possibility! if! the! node! could! run! on! a! 2000! mAh! battery!for!at!least!a!week,!while!obtaining!a!reading!from!each!sensor!at!least!every!hour.!
The! wireless! communication! is! part! of! Aji! John! Pushpam’s! thesis! and! will! not! be! included! in! this! thesis.! Similarly,!sensor!data!will!be!analysed!only!for!its!reliability!and!accuracy:!trend!analysis!and!evaluation!of! the!actual!air!quality!is!also!part!of!Aji!John!Pushpam’s!thesis!and!will!be!excluded!from!this!thesis. !
1.4#Project#stakeholders#
Four!main!groups!of!stakeholders!were!identified,!which!are!either!directly!involved!in!the!development!of! the!sensor!node!or!are!affected!by!the!air!quality!monitoring.! !• The! primary! stakeholder! is! Uppsala! Kommun.! Two! of! the! three! sensor! nodes! are! located! in! their! facilities.! They! will! benefit! from! the! knowledge! of! air! quality! gathered! during! the! research.! The! findings!of!this!research!can!help!create!further!initiatives!to!improve!air!quality!on!a!citywide!level.! For!example,!locations!for!new!infrastructure!and!construction!work!can!be!chosen!in!areas!where! air!pollution!is!low.!!Additionally!the!sensor!data!will!provide!important!statistical!information,!which! can! relate! to! other! projects! and! span! future! research! topics,! such! as! the! influence! of! traffic! and! construction!on!air!quality.! ! • Uppsala!University!is!a!stakeholder!from!an!educational!point!of!view.!They!provide!technical!and! scientific!advice,!as!well!as!guidance!during!the!execution!of!the!project,!and!an!assessment!based! on!the!graduation!report!and!presentation.!The!research!is!also!funded!by!Uppsala!University.! !
• Uppsala! Stadsteater! is! a! stakeholder! since! they! are! offering! their! facilities! for! sensor! node! installation.! They! are! not! interested! in! city/wide! air! quality! measurements! per! se,! but! because! of! their!central!location!and!large!amount!of!daily!visitors,!air/quality!is!a!concern!that!also!affects!them! directly.!
!
• The! stakeholders! that! are! affected! by! the! consequences! of! the! project! are! inhabitants! of! Uppsala! city.!This!research!is!a!step!towards!publically!available!air!quality!data,!which!will!allow!individuals! to! create! initiatives! through! Uppsala! Kommun! to! improve! the! city’s! infrastructure! and! therefore! mitigate!air!pollution,!improving!the!quality!of!life!of!the!city’s!residents.!
2%%Situational%analysis%and%related%work
Some!statistical!information!on!Uppsala!and!Sweden!will!be!discussed,!as!well!as!some!existing!solutions!for! air!quality!monitoring!that!have!been!implemented.!The!measurement!parameters!that!are!important!for!air! quality!will!be!discussed,!as!well!as!the!measurement!principles!used!in!different!sensors.!2.1#Situational#Analysis#
Sweden!currently!has!a!total!population!of!about!9.8!million!people,!and!the!population!of!Uppsala!is!about! 350!thousand!as!of!the!first!half!of!the!year!2015.![1].!The!population!of!Uppsala!has!grown!by!a!full!1%!in! the!last!half/year![2],!as!compared!to!the!statistics!of!the!end!of!year!2014.! Statistics!Sweden!(SCB),!which!is!Sweden’s!central!statistics!bureau,!predicts!that!in!the!next!ten!years!(by! 2025),!the!population!of!Sweden!will!increase!by!one!million![3]. Regarding!traffic!data,!the!number!of!newly!registered!vehicles!in!Sweden!in!September!2015!has!seen!an! increase!of!11.8%,!compared!to!the!same!period!last!year,!according!to!SCB![5].! A!breakdown!of!the!kind!of!vehicles!that!were!registered!in!Uppsala!in!2015!can!be!seen!in!Table!1![6]:! ! Table!1:!Newly!registered!vehicles!in!Uppsala!by!type,!2015As!Table!1!shows,!electrical!or!hybrid!vehicles!are!still!a!novelty,!and!there!is!a!steady!amount!of!diesel!and! gas! vehicles! being! registered! every! month.! The! large! amount! of! gas! vehicles! are! a! major! source! of! air! pollution,!and!according!to!these!statistics,!electrical!vehicles!constitute!a!very!small!part!of!newly!registered! cars.! Both! the! large! increase! in! population! and! the! large! proportion! of! gas! and! diesel! powered! vehicles! highlight!the!importance!of!air!quality!monitoring!and!providing!the!data!to!the!public.!
Statistics! Sweden! does! not! provide! detailed! information! on! air! quality! as! such! research! has! not! been! done! nation/wide.! This! is! indicated!by!public!statistics!being!scarcely!available!and!for!some! parameters!non/existent.!There!has!been!detailed!research!on!the! increase! of! CO2! levels,! but! this! parameter! gives! very! limited!
information,! as! CO2! is! only! harmful! to! humans! in! very! high!
concentrations,!and!other!gasses!such!as!the!ones!measured!in!our! research,! are! a! much! bigger! threat,! but! have! not! yet! been! investigated![4].
The!Uppsala!Kommun!and!the!Swedish!Environmental!Protection!Agency!own!two!particle!measuring!units! in!Uppsala,!which!are!located!at!two!different!locations:!one!at!Klostergatan!and!one!at!Kungsgatan![8].!The! mass!of!particles!in!the!air!is!measured,!so!dust!and!other!solid!particles!such!as!allergens!can!be!quantified.! Humidity,! temperature! and! NO2! are! also! measured.! In! other! cities! such! as! Stockholm! similar! units!
additionally!measure!O3!concentrations.!There!are!a!total!of!12!such!air!quality!measuring!units!in!Sweden!
with!publicly!available!data![9].!This!is!less!than!in!Norway!or!Finland,!and!countries!in!central!Europe!also! have!far!more!detailed!and!distributed!data!of!air!quality!available,!as!can!be!seen!in!Figure!1.!The!high!cost! and!low!coverage!of!the!equipment!used!in!Sweden!highlight!the!need!for!a!low/cost!distributed!solution. There!is!also!a!Vinnova!funded!project!initiated!by!Uppsala!University!–!Green!IoT!–!with!the!participation!of! several!companies,!which!has!started!in!September!2015![7].!The!aim!of!the!project!similarly!is!to!develop!an! air!quality!monitoring!network.!The!project!is!currently!in!the!stages!of!developing!the!sensor!nodes,!and!is!a! two/year!project.! ! Figure!1:!Air!quality!measurement!locations!in!Europe![9]
2.2#Background#on#air#quality#measurement#
The! United! States! Environmental! Protection! Agency! (EPA)! suggests! 6! parameters! that! provide! a! complete! overview! of! the! air! quality! [10].! These! are:! ozone,! particulate! matter,! carbon! monoxide,! nitrogen! oxide,! sulphur!oxide!and!lead.!The!EPA!usually!presents!the!sensor!readings!as!scaled!values:!a!number!from!0!to! 500! and! this! is! known! as! the! Air! Quality! Index! (AQI).! The! scaling! is! done! to! obtain! a! value! that! is! more! understandable! to! the! general! public.! The! scaling! is! done! differently! for! each! parameter! and! based! on! thresholds!that!are!considered!safe!or!unsafe!for!health.!These!standards!are!also!subject!to!change!and!are! described!in!more!detail!on!the!EPA’s!website.! AQI!values!from!0!to!50!represent!good!air!quality,!50!to!100!moderate,!an!AQI!above!150!being!unhealthy! and!above!300!being!hazardous![11].!In!this!thesis,!the!sensor!readings!will!be!presented!in!actual!scientific! units!instead,!and!conversion!to!an!AQI!index!will!be!excluded!from!the!scope!of!the!report,!since!the!focus!is! on!sensor!accuracy!and!performance,!instead!of!interpreting!the!measurements.! 4!of!the!6!suggested!parameters!were!chosen!for!the!scope!of!this!thesis:!particulate!matter, ozone,!carbon! monoxide! and! nitrogen! oxide.! Lead! and! sulphur! oxide! were! excluded,! as! these! occur! comparatively! more! rarely!and!mostly!near!production!plants!and!industrial!areas![11].!Additionally,!humidity!and!temperature! are!also!monitored!as!these!give!useful!additional!information!for!trend!analysis!and!calibration.!
When!considering!gas!sensors,!there!are!two!main!types!available:!electrochemical!and!semiconductor!metal! oxide! (SMO)! sensors.! The! main! difference! is! the! operating! principle:! electrochemical! sensors! generate! a! current!proportional!to!the!gas!concentration!and!therefore!require!less!power,!whereas!SMO!sensors!are! easier!to!integrate!but!require!heating!for!the!chemical!reaction!to!occur,!therefore!requiring!more!power! [13].!Since!the!focus!of!this!thesis!is!not!on!power!saving,!both!types!of!sensors!can!be!considered.!
Further! on,! other! similar! research,! using! various! sensing! methods! for! the! chosen! parameters,! will! be! discussed,! evaluating! advantages! and! drawbacks! while! keeping! in! mind! the! low/cost! requirement.! This! evaluation!will!then!be!the!basis!for!choosing!the!most!suitable!sensors!for!the!application!in!this!research.!
2.3#Related#work#
Similar!research!has!been!done!on!creating!low/cost!compact!air!quality!monitoring!solutions!and!networks.! Some!of!the!notable!examples!will!be!introduced!here!and!the!differences!and!similarities!with!this!research! will!be!discussed.! Some!other!recently!conducted!research,!by!D.!Hasenfratz!et.!al.!in!2015![12]!and!by!P.!Zappi!et.!al.!in!2012! [13]!also!choose!to!monitor!these!same!4!main!parameters.!Both!of!these!research!papers!similarly!aim!to! develop!low/cost!compact!solutions!that!could!possibly!replace!the!nowadays!more!common!expensive!and! large!industrial!equipment.!They!also!accentuate!the!need!for!more!widespread!measurement!locations!and! easily!accessible!public!data,!which!is!similar!to!the!research!in!this!thesis.!The! main! difference! with! the! research! done! by! D.! Hasenfratz! et.! al.! is! that! their! goal! is! to! also! create! a! mobile!network!which!is!attached!to!public!transport,!whereas!this!research!focuses!only!on!static!nodes.! Some!of!the!sensors!they!use!are!also!not!suitable!for!this!research!due!to!their!high!cost,!for!example,!the! Alphasense! CO/B4! sensor,! which! is! priced! at! around! 1500! SEK.! In! the! paper,! it! is! also! discussed! that! for! mobile!nodes,!a!specially!designed!protection!cover!is!necessary!to!protect!the!sensors!from!direct!varying! airflow!that!could!affect!the!reading!accuracy.!It!is!also!noted,!however,!that!for!conditions!where!airflow!is! not!rapidly!changing,!such!as!this!research,!a!special!cover!is!not!necessary.! Another!difference!is!that!the!nodes!in!this!research!are!not!equipped!with!an!air!fan,!whereas!the!research! of!D.!Hasenfratz!et.!al.!uses!a!fan!to!draw!air!out.!It!is!said!that!this!is!done!to!achieve!more!stability.!The! influence!of!airflow!and!the!necessity!of!a!fan!are!however!not!discussed!in!any!detail,!so!it!will!be!referred! to!other!papers!to!investigate!the!effects.! The!research!by!P.!Zappi!et.!al.!similarly!to!this!research!does!not!use!a!fan!for!airflow!but!instead!relies!on! openings!in!the!casing!to!ensure!air!exchange.!The!focus!of!this!thesis!is!also!similarly!on!developing!low/cost!
nodes,!with!a!focus!on!low!power!consumption!as!well.!The!authors!claim!to!obtain!good!results!for!both! static! and! wearable! sensor! nodes! without! the! use! of! an! additional! fan! for! airflow.! The! choice! can! be! motivated!by!both!extra!cost!and!power!consumption!involved!with!adding!an!air!fan.!
A!research!by!Y.!Jiang!et.!al.!which!focuses!on!developing!sensor!nodes!for!indoor!CO2!monitoring!discusses!
the!benefits!of!adding!an!air!fan!in!more!detail![35].!It!is!said!that!a!fan!that!gives!an!air!exchange!of!about!2! l/min! is! used! to! ensure! there! is! no! latency! in! the! sensor! readings.! So! in! the! case! of! measuring! gas! concentration,!this!implies!that!without!a!fan,!the!sensor!readings!may!not!reflect!the!current!concentration,! but!be!delayed!and!therefore!also!less!sensitive!to!sudden!changes.!This!statement!can!be!verified!by!this! thesis!since!a!fan!will!not!be!used.!
Research! by! J.! H.! Liu! et.! al.! similarly! develops! a! stationary! node! for! air! quality! monitoring! in! urban! environments![36].!Instead!of!using!a!fan,!they!make!an!opening!on!the!side!of!the!sensor!casing!and!cover!it! using!a!waterproof!breathing!membrane.!They!then!compared!their!measurements!of!CO!concentrations!to! those! of! local! industrial! monitoring! stations.! With! nine! nodes,! the! correlation! coefficient! R2! was! 0.67! at!
lowest!and!0.86!at!highest.!These!results!suggest!that!good!sensor!values!can!be!obtained!that!are!highly! comparable!to!industrial!measurements!even!without!the!use!of!a!fan!for!airflow.!This!research!will!similarly! make!an!opening!in!the!casing!for!airflow,!however,!a!waterproof!membrane!will!not!be!used!as!this!would! interfere! with! the! particle! concentration! readings,! since! larger! particles! may! not! pass! through! the! membrane.
Research!conducted!by!S.!Choi!et.!al.!also!aims!to!develop!both!the!hardware!and!the!software!for!compact! and! low/cost! air! quality! sensor! nodes! [37].! They! however! in! their! conclusions! argue! for! the! complete! opposite:!that!small!and!low/cost!air!quality!sensors!have!too!much!manufacturing!variability!and!that!the! readings!between!sensors!are!too!different!to!give!an!accurate!value.!It!is!suggested!that!recent!advances!in! technology,!such!as!the!miniature!MEMS!gas!sensors,!help!make!the!sensor!nodes!more!compact,!but!do!not! provide!enough!accuracy!to!match!industrial!sensors.!
Research!by!N.!Kularatna!et.!al.!successfully!design!a!sensor!node!intended!for!outdoor!measurements!using! low/cost! semiconductor! gas! sensors! [38].! The! sensors! are! lab/calibrated,! as! opposed! to! this! thesis! where! field! calibration! will! be! performed.! However,! their! results! suggest! that! modern! low/cost! gas! sensors! are! highly!suitable!and!can!even!meet!international!IEEE!standards,!according!to!results!in!a!lab!environment.! There! therefore! exist! conflicting! research! evidence! on! both! the! mechanical! design! (whether! an! air! fan! is! necessary)!and!on!whether!accurate!results!can!be!achieved!using!low/cost!gas!sensors.!This!research!can! hopefully!add!to!the!already!existing!knowledge!and!obtain!results!supporting!either!of!these!opinions.!
3%–%Conceptual%model
!The!problem!of!measuring!air!quality!is!causal.!This!is!due!to!there!already!being!literature!available!about! air!quality!measurement,!and!a!solution!for!this!purpose!is!developed!in!this!thesis.!Information!on!which! parameters!are!important!for!measuring!air!quality!and!what!requirements!should!be!met!are!given!in!the! previous! sections.! In! this! section,! a! particular! sensor! choice! will! be! made,! which! should! match! the! requirements! defined! in! the! Rationale.! The! two! main! requirements! are! accuracy! and! cost.! Cost/wise,! the! majority! of! the! budget! is! chosen! to! be! spent! on! the! gas! sensors! and! the! particle! sensor,! which! are! most! important!for!this!research.!The!microcontroller!and!humidity!and!temperature!sensors!are!secondary,!and! many!low/cost!alternatives!are!available!so!budget!will!be!saved!on!these.!
Accuracy! is! also! important.! Therefore! a! theoretical! calculation! of! the! sensor! response! will! be! made! when! possible!to!ensure!that!the!chosen!sensor!is!suitable.! !
3.1#Microcontroller#
! A!Seeeduino!Stalker!board!was!chosen!for!the!prototype!node![14].!It!uses!an!ATMega168!microcontroller! [15]!and!has!several!added!peripherals!for!easy!prototyping.!The!main!factors!considered!were!the!ease!of! programming!of!the!prototype!using!the!Arduino!interface,!sufficient!amount!of!pins!for!sensors!(15!digital,!6! analog!pins),!integrated!I2C!communication!which!allows!the!use!of!a!broader!range!of!possible!sensors,!and! the!interface!for!directly!connecting!an!XBee!radio.!The!radio!communication!part!is!discussed!in!Aji!John! Pushpam’s!thesis.! The!board!has!a!10/bit!analog!to!digital!converter!(ADC),!which!means!1023!conversion!steps!(excluding!0).!It! is!also!possible!to!switch!between!two!measurement!ranges:!0!to!5!volts!and!0!to!1.1!volts.!These!will!be! used!further!on!when!interfacing!the!sensors.!Power! consumption! was! not! an! important! factor! in! choosing! the! microcontroller,! since! the! sensors! are! expected!to!use!much!more!power!in!comparison!(for!example,!due!to!the!heating!principle!for!gas!sensors).! A!more!power/efficient!microcontroller!can!be!chosen!for!future!research,!and!ease!of!implementation!was! the!primary!factor!for!the!scope!of!this!thesis.! The!typical!expected!current!consumption!is!given!as!300mA!in!the!datasheet,!however,!this!is!most!likely!an! estimate!for!when!the!other!interfaces!(such!as!the!radio)!are!used!and!is!too!high!to!be!the!current!draw!for! just!the!board.!An!estimation!based!on!a!measurement!during!performance!is!given!on!the!manufacturer’s! website! [16],! stating! between! 3.6! and! 22.4mA! during! active! mode.! The! actual! power! draw! for! our! application!will!therefore!have!to!be!measured!and!will!be!given!in!the!Results!section.!
3.2#Temperature#and#humidity#sensor#
!
Two!different!humidity!and!temperature!sensors!that!met!the!measurement!range!requirements!and!were! both! low/cost! were! considered.! The! first! option! considered! was! ChipCap! 2! sensor! by! General! Electronics! [19],! which! is! a! joint! temperature! and! humidity! sensor.! The! second! option! considered! was! an! Epcos! B57020m2502a17!temperature!sensor![22]!and!a!Multicomp!hcz/h8/b!humidity!sensor![23].!
The!advantages!of!the!second!option!considered!is!that!both!of!the!sensors!are!rugged!and!can!withstand! various! environmental! conditions,! especially! the! temperature! sensor,! which! is! also! waterproof.! However,! both! of! these! sensors! give! an! analog! output! (by! changing! resistance! based! on! the! measured! conditions).! Analog! sensors! are! simple! to! implement,! but! are! intrinsically! more! prone! to! noise! and! inaccuracy! when! converting!the!analog!reading.!
The!ChipCap!2!sensor!on!the!other!hand!uses!digital!communication!(I2C!protocol)!and!both!the!humidity!and!
temperature!sensors!are!pre/calibrated.!It!has!excellent!accuracy!characteristics!and!a!wide!measurement! range!that!even!exceeds!the!defined!requirements.!The!temperature!measurement!has!±0.3!ºC!error!in!the!
20!to!40ºC!range!and!no!more!than!±2.0!ºC!error!in!the!full!/40!to!+125!ºC!temperature!range.!The!humidity! measurement!has!±2%!error!in!the!20!to!80%!relative!humidity!range!and!no!more!than!±4%!error!in!the!full! range.! The!large!measurement!range!ensures!that!the!sensor!continues!to!be!operational!in!cold!winters.!This!is! especially!important!because!in!low!temperatures,!the!accuracy!and!performance!of!the!other!sensors!may! suffer,!and!this!will!be!known!because!of!the!accurate!temperature!reference.!The!sensor!is!not!waterproof! as! compared! to! the! other! option,! but! weatherproofing! will! be! achieved! by! the! mechanical! design! of! the! casing!instead.!
3.3#Particle#concentration#sensor#
! Two!low/cost!particle!concentration!sensors!were!considered!for!the!sensor!node:!the!Sharp!GP2Y1010AU0F! [17]!and!Shinyei!PPD42NS![18].!The!sensors!differ!in!the!sense!that!the!Sharp!sensor!measures!the!mass!of! the!total!particles!in!a!volume,!while!the!Shinyei!sensor!measures!particle!count!of!all!particles!greater!than! 1um!in!size.!It!has!been!suggested!by!research!(A.!Tittarelli!et.!al.![20],!Wichmann!et.!al.![21])!that!particle! count! is! the! leading! cause! of! health! problems,! instead! of! particle! mass.! Particle! mass! is! the! most! widely! measured! parameter! though,! as! sensors! for! measuring! particle! mass! have! historically! been! more! widely! available![29].!It!was!chosen!to!measure!particle!count!using!the!Shinyei!sensor,!since!particle!count!has!a! much!more!direct!correlation!to!health!risks.!The!Shinyei!sensor!has!a!higher!current!consumption!(90mA!as! compared!to!20mA!for!the!Sharp!sensor),!because!of!it!using!an!infrared!laser!in!its!measurement!principle.! Power!consumption!however!is!not!the!main!concern!for!this!research.! Furthermore,!it!could!be!a!possibility!to!convert!particle!count!to!total!particle!mass!(to!make!readings!of! both!sensors!comparable).!It!has!been!suggested!by!A.!Tittarelli!et.!al.![20]!that!there!is!strong!correlation! between! particle! count! and! particle! mass! (correlation! coefficient! R2=0.734! for! particles! of! size! 0.3um! to!10um!and!above).!However!the!correlation!is!not!strong!enough!to!make!a!direct!conversion!and!according! to!the!authors!requires!more!research.!It!is!however!a!good!enough!correlation!to!make!trend!comparison! and! analysis,! so! the! particle! count! readings! will! be! comparable! with! other! sensors! that! measure! particle! mass.
The! Shinyei! particle! sensor! is! factory/calibrated! and! uses! a! measurement! principle! of! pulse/width! modulation.! During! a! 30! second! period,! the! total! duration! of! high/pulses! is! measured! and! this! is! then! expressed!as!a!0!to!1!ratio.!The!conversion!formula!obtained!from!the!curve!in!the!datasheet![18]!is!given! below!in!Equation!1:!
Ratio = 100 × (High pulse duration / 30 seconds) Output = 1.1× Ratio3
- 3.8 × Ratio2
+ 520 × Ratio + 0.62! Equation!1:!Reading!conversion!formula!for!Shinyei!particle!sensor
The! converted! output! is! in! pieces! per! 283ml,! which! can! then! further! be! scaled! to! other! units! such! as! pieces/liter.!
SGX! Sensortech:! MiCS/2614! ozone! sensor! [24]! and! MiCS/2714! nitrogen! dioxide! sensor! [25].! Both! sensors! function!as!heated!resistors,!whose!resistance!varies!with!the!gas!measured!due!to!a!chemical!reaction!on! their! surface.! Both! sensors! have! very! good! sensitivity! characteristics,! as! the! measurement! ranges! are! 10/ 1000ppb!for!ozone and 0.05/10ppm!for!nitrogen!dioxide.!Both!sensors!are!said!by!the!manufacturer!to!be! designed!for!harsh!environments.!
Both! sensors! need! specific! resistor! values! to! achieve! the! right! heating! temperature.! According! to! the! datasheets,!these!values!are!131!ohms!for!the!ozone!sensor!and!82!ohms!for!the!nitrogen!dioxide!sensor.! There! are! no! standard! resistor! values! that! match! these! values,! so! resistors! were! combined! in! series! and! parallel!to!obtain!a!value!close!to!the!desired. For!the!ozone!sensor,!a!100!and!47 ohm!resistors!in!parallel!plus!a!100!ohm!resistor!in!series!was!used!to! achieve!a!resistance!value!of!about!131.9!ohms!(the!aim!was!132!ohms).!For!the!nitrogen!dioxide!sensor,! two!resistors!in!parallel!of!100!and!470!ohms!were!used!to!achieve!a!value!of!82.4!ohms!(the!aim!was!82! ohms).!The!calculations!are!given!in!APPENDIX!D. Further!on,!the!sensor!requires!a!voltage!divider!to!convert!the!sensor’s!resistance!to!a!voltage.!The!voltage! is! then! measured! using! the! ADC (analog to digital converter)! of! the! microcontroller,! and! this! is! converted! back!to!the!variable!resistance!of!the!sensor,!which!is!converted!to!gas!concentration!later.!The!formula!for! calculating!the!variable!resistance!based!on!output!voltage!of!the!voltage!divider!is!given!in!Equation!2![26].! !
R
1=
R
2⋅V
inV
out− R
2= R
2⋅ (
V
inV
out−1)
! Equation!2:!Formula!of!voltage!divider!output, Vout!is!the!measured!output!and!R1!is!the!varying!resistance!of!the!sensor!Vin!is!a!5!volt!input,!Vout!is!the!output!voltage!measured!by!the!ADC,!R2!is!a!known!resistance!value!for!scaling,!
and!R1!is!the!unknown!sensor!resistance!obtained!as!a!result!of!the!calculation.!
According!to!the!datasheets!of!the!sensors,!the!R1!value!can!have!many!different!ranges!depending!on!the!
manufacture! variability! of! the! sensors [24]! [25].! For! the! ozone! sensor,! the! typical! resistance! in! air! is! 11! Kohms! but! can! vary! from! 3! to! 60! Kohms,! and! for! nitrogen! dioxide,! it! can! vary! from! 0.8! to! 20! Kohms.! Therefore!it!is!difficult!to!select!a!suitable!R2!value!for!all!sensors!since!the!possible!range!of!R1!is!different!
for! each! batch! of! sensors.! Furthermore,! to! actually! test! what! the! resistance! in! air! is,! each! sensor! would! require!calibration!in!a!cleanroom!using!synthetic!air,!which!is!not!possible!in!the!scope!of!this!project.!It!is! recommended!in!the!datasheets!that!R2!should!not!be!smaller!than!820!ohms,!since!this!can!damage!the!
sensor [24]![25].!A!value!of!1!Kohm!was!chosen!for!R2!and!it!will!be!verified!experimentally!whether!this!is!
suitable!for!all!sensors!and!what!is!the!resulting!resolution!/!accuracy.!
Once!R1!in!ohms!has!been!obtained!from!the!voltage!measurement!conversion,!it!can!be!further!converted!
to! an! actual! gas! concentration! value! in! parts! per! billion! (ppb).! The! conversion! formula! to! convert! the! measured! resistance! readings! to! actual! units! was! only! given! in! graphs! by! the! manufacturer.! Therefore! another!graph!was!overlaid!on!top!of!the!image!and!a!polynomial!mapped!through!the!points!to!in!the!end! obtain!a!conversion!formula.!Also,!the!x!and!y!axes!had!to!be!flipped!as!compared!to!the!original!graph!of!the! datasheet,!so!that!now!the!x!axis!is!a!measured!resistance!ratio!and!the!y!axis!is!the!gas!concentration.!The! exact!polynomials!are!given!in!APPENDIX!A.
The!obtained!calibration!curves!and!their!fitting!polynomials!are!given!below!in!Figure!2!and!Figure!3. !
! Figure!2:!Ozone!sensor!calibration!curve
! Figure!3:!Nitrogen!dioxide!calibration!curve
Rs!in!the!graph!labels!is!the!same!as!R1!from!Equation!2.!The!extra!dividing!factors!for!the!x!axis!(R100ppb!and!R0)!
are!the!sensor’s!resistance!at!a!known!gas!concentration.!It!is!simply!a!scaling!factor!for!R1!that!is!applied! before! it! is! passed! into! the! polynomial! as! the! x! value.! This! scaling! factor! will! be! obtained! by! comparing! readings! to! a! pre/calibrated! commercial! sensor! node! (Air! Quality! Egg).! It! is! not! necessary! in! this! case! to!
3.5#Carbon#monoxide#sensor#
!
For! measuring! carbon! monoxide,! two! various! sensors! were! considered.! The! first! one! was! MiCS/5121WP! CO/VOC! Sensor! [27]! and! the! second! was! SGX/4CO! Industrial! Carbon! Monoxide! Sensor! [28],! both! manufactured!by!the!company!SGX!Sensortech.! Both!sensors!have!their!advantages!and!disadvantages.!The!CO/VOC!sensor,!as!the!name!suggests,!measures! both!carbon!monoxide!and!VOC!(volatile!organic!compounds).!VOCs!are!usually!produced!during!burning!of! organic!substances!or!from!pollen!that!can!cause!allergies!and!therefore!is!also!an!interesting!parameter!to! measure.!This!sensor!is!also!more!compact!and!easier!to!implement:!it!is!a!resistive!type!sensor!(similarly!to! the!previously!discussed!gas!sensors)!and!requires!only!a!simple!resistor!bridge.! The!other!sensor!(SGX/4CO)!is!larger,!more!expensive!and!difficult!to!implement!because!it!requires!an!extra! amplifier!chip!to!condition!its!output.!This!sensor,!however,!offers!industrial/grade!accuracy!and!has!a!linear! response!to!the!gas!concentration.!Furthermore,!it!is!much!more!selective!and!VOCs!are!not!detected!with! this!sensor.!Since!reacting!to!two!different!substances!could!lead!to!false!positives,!it!was!considered!much! more!valuable!to!have!a!carbon!monoxide!reading!without!the!VOC!influence.!The!second!option!(SGX/4CO)! was!therefore!chosen.! The!SGX/4CO!gives!a!current!output,!which!needs!to!be!converted!to!a!voltage!to!be!measured.!To!achieve! this,!a!transimpedance!amplifier!is!used![29].!The!schematics!will!be!presented!further!on!in!the!report.! The!optimal!amplification!factor!of!the!amplifier!needs!to!be!calculated.!The!sensor!gives!a!response!of!70!±! 25nA/ppm.!We!first!assume!maximum!sensitivity!(95nA/ppm)!to!avoid!saturation!at!the!maximum!desired! CO!concentration.!The!maximum!concentration!range!of!interest!is!9!ppm,!since,!according!to!the!Clean!Air! Act,!an!adult!should!not!be!exposed!to!this!concentration!for!more!than!8!hours!a!maximum!of!3!times!per! year.!We!switch!to!the!internal!1.1V!ADC!voltage!reference!of!the!Seeeduino!Stalker!board!to!obtain!better! resolution,!since!very!small!currents!and!voltages!are!dealt!with.! Using!the!formulas!provided!in!the!application!notes!document!for!this!sensor![31],!it!is!possible!to!calculate! the!necessary!amplification.! !
Sensor sensitivity = 95 nA/ppm = 95⋅10
-9A/ppm
Equation!3:!Maximum!sensitivity!of!CO!sensor!
The! measurement! range! we! aim! for! is! 0/9ppm,! 9ppm! then! equaling! the! maximum! of! the! 1.1V! ADC! measurement!range!(and!1023!ADC!steps):! !
System sensitivity = 1.1V / 9ppm = 0.122V / ppm
Equation!4:!Desired!sensitivity!of!CO!sensor! This!is!the!system!sensitivity!that!we!aim!for,!and!this!allows!to!calculate!the!necessary!amplification!using! the!following!formula:! !System sensitivity = Transimpedance Gain ⋅ Sensor Sensitivity
Transimpedance Gain = System sensitivity / Sensor Sensitivity = 1284210
Equation!5:!Required!transimpedance!gain!for!CO!sensor!
The!amplification!should!be!very!high,!in!the!order!of!a!million!times.!The!closest!standard!resistor!value!is! chosen,!which!is!a!1!Mohm!resistor.!We!then!recalculate!to!obtain!what!the!actual!values!will!be!using!this! approximate!resistor!value.!The!range!of!measurement!using!a!1!Mohm!resistor!can!be!calculated!as:!
!
System sensitivity = Transimpedance Gain ⋅ Sensor Sensitivity = 1000000 ⋅ 95⋅10
-9= 0.095 V/ppm
!
This!means!we!have!a!resolution!(sensitivity)!of!approximately!0.1V!per!ppm!(100!ADC!steps!per!ppm).!The! maximum!measurement!range!is:!
1.1V / System sensitivity = 1.1 / 0.095 = 11.58 ppm
Equation!7:!Maximum!measurement!range!of!CO!using!1.1V!reference However,!the!maximum!measurement!range!can!be!increased!by!switching!the!ADC!voltage!reference!to!5V,! in!which!case!we!obtain:5V / System sensitivity = 5 / 0.095 = 52.63 ppm
Equation!8: Maximum!measurement!range!of!CO!using!5V!reference The!sensor!can!therefore!theoretically!operate!in!two!modes:!up!to!approx.!11.6!ppm!with!larger!resolution! or!up!to!approx.!52.6!ppm!with!less!resolution.!However,!the!CO!concentrations!are!not!expected!to!exceed! 9!ppm!(which!is!already!a!hazardous!value),!therefore!the!lower!measurement!range!(0/11.6!ppm)!with!more! resolution!is!the!one!that!will!be!used.! !3.6#Cost#estimation
The!estimated!costs!per!sensor!node!are!presented!in!Table!2.!!The!conversion!from!USD!to!SEK!was!done! using!the!conversion!rate!provided!by!Google!on!20.01.2016.!The!total!cost!is!below!the!set!goal!of!2000!SEK.!!
Table!2:!Cost!estimation!and!parts!list!per!sensor!node!
Part! Notes! Cost!(USD)! Cost!(SEK)
Humidity!and!
temperature!sensor! CC2D33S!Bulk!ChipCap2,!digital,!sleep!mode,!3%,!3.3v! 6,00! 51,29! Ozone!sensor! SGX!Sensortech!Limited!MICS/2614! 8,10! 69,24! Nitrogen!dioxide!
sensor! SGX!Sensortech!Limited!MICS/2714!! 8,10! 69,24! Carbon!monoxide!
sensor! SGX/4CO!Industrial!Carbon!Monoxide!Sensor!Application!:!Portable!Gas! Detectors! 36,33! 310,57! Particle!sensor! Shinyei!Particle!Sensor!(PPD42NS)! 15,90! 135,92! Seeeduino!Stalker! v3!board! ATMEL!/!ATMEGA328!microcontroller! 7,90! 67,53! XBee!radio! XBEE!PRO!868MHZ!1/315MW!RPSMA!EU! 60,12! 513,93! Casing! Universal/Gehäuse!Polycarbonat!Blau!150! x!80!x!50!Hammond!Electronics!1591! DTBU! 10,00! 85,48! Overhead!expenses! An!estimated!25%!of!total!cost,!including! extra!materials!such!as!wiring,!glue! 38,11! 325,80! Total! !! 190,56! 1629,02!
4%–%Research%design%
For!this!research,!central!locations!in!the!city!were!preferred!in!order!to!be!able!to!measure!more!prominent! pollution.! There! is! an! existing! air! quality! measurement! unit! on! Kungsgatan! with! which! results! could! be! compared.! The! ideal! location! would! therefore! be! on! Kungsgatan! and! near! the! measurement! unit.! The! planning!of!testing!is!further!described!and!the!results!of!these!experiments!are!presented!in!the!Research! Results!section.!
!
4.1#Measurement#location#and#setup#
Three! locations! in! the! city! were! chosen! for! the! three! sensor! nodes,! creating! a! small! air! quality! sensor! network.! The! locations! were! all! in! the! center! of! Uppsala! and! near! one! of! the! busiest! streets! traffic/wise:! Kungsgatan.!This!choice!of!the!location!of!measurements!was!based!on!two!main!considerations:! 1. It!is!expected!to!see!more!variation!in!air!pollution!near!busy!traffic! 2. There!is!an!air!quality!measurement!unit!on!Kungsgatan!(owned!by!the!Kommun)! ! For!this!research,!it!is!more!interesting!to!focus!on!central!areas!of!the!city!where!air!pollution!is!expected!to! be!more!prominent.!Furthermore,!the!existing!air!quality!measurement!unit!(which!was!briefly!mentioned!in! the!Introduction)!can!provide!reference!data!for!comparison.!All!three!node!locations!are!close!to!this!unit,! in!a!radius!of!less!than!100!meters.!
Two! nodes! were! planned! to! be! placed! at! the! Uppsala! Kommun,! on! the! second! floor! of! the! building.!One! node! was! planned! to! be! placed! at! Uppsala! Stadsteater! on! the! first! floor! balcony.!The! outer! casing! of! the! nodes!would!be!fastened!/!glued!near!the!wall!of!each!building!outside!a!window,!and!the!power!cable!of! the!node!would!be!dragged!inside!the!building!through!the!window.!The!sensor!nodes!are!powered!using! mains! power,! and! since! there! are! power! sockets! available! indoors,! this! provides! weatherproofing.! In! the! case!of!Uppsala!Stadsteater,!a!power!socket!outdoors!is!provided!by!them!that!is!weatherproof.!The!sensor! nodes! should! be! placed! facing! so! that! the! opening! for! air! is! pointing! downwards,! so! to! avoid! damage! to! electronics!from!rain!and!moisture.!
4.2#Measurement#collection,#calibration#and#analysis#
Two! separate! gateways! are! used for the Kommun! building! and! Stadsteater,! one! in! each! building.! Sensor! readings!are!sampled!approx.!every!3!minutes!and!then!sent!to!the!gateway!using!the!868!MHz!XBee!radio.! Normally! for! measuring! air! quality,! some! parameters! can! be! sampled! once! a! day! or! even! more! rarely.! However,!it!is!interesting!for!research!to!collect!samples!more!often!in!order!to!see!trends!in!the!readings! and!later!re/evaluate!on!how!often!sampling!should!be!done.!The!gateway!forwards!the!readings!to!a!Cloud! database,!where!they!can!be!observed!real/time!and!further!analyzed.!The!radio!communication!and!data! collection!are!discussed!in!detail!in!Aji!John!Pushpam’s!thesis.!
The! sensor! readings! will! be! calibrated! using! the! reference! method! Air! Quality! Egg.! Only! the! gas! sensor! accuracy!and!readings!will!be!discussed,!as!these!are!the!only!ones!requiring!calibration.!The!readings!will!be! calibrated!after!being!received!on!the!Cloud!database;!this!is!done!in!order!to!first!be!able!to!observe!the! raw! readings! sent! by! the! sensor! node.! Calibration! will! be! done! by! taking! a! measurement! by! both! the! Air! Quality!Egg!and!the!sensor!node!in!the!same!location!and!conditions!and!scaling!the!sensor!node’s!readings! to!match!those!of!the!Air!Quality!Egg.!!After!having!been!used!as!a!calibration!reference,!the!Air!Quality!Egg! will!be!used!as!a!reference!for!stability!and!accuracy,!by!comparing!all!further!readings!after!calibration.!
The!main!method!used!for!evaluation!of!the!sensor!data!will!be!the!Pearson’s!correlation!coefficient!r![33].! The!correlation!coefficient!is!a!number!between!0!and!1!describing!how!well!data!follows!the!trends!of!some! reference!data.!1!means!perfect!correlation!and!0!means!no!correlation.!Negative!correlation!coefficient!(/1! to!0)!means!that!the!correlation!is!inverted:!when!the!data!increases,!the!reference!value!decreases,!and! vice!versa.!
The! sensor! data! will! first! be! plotted! in! comparison! to! the! reference! Air! Quality! Egg.! Then! the! correlation! coefficients!will!be!calculated!between!the!node’s!data!and!temperature,!humidity!and!Air!Quality!Egg!data.! Further!on!temperature!compensation!may!be!applied!to!improve!correlation.!In!a!field/calibration!as!in!this! research,!temperature!calibration!is!difficult!to!perform!because!none!of!the!variables!measured!are!kept! constant!(as!opposed!to!a!lab!environment).!Information!on!temperature!compensation!for!the!gas!sensors! is!not!available,!and!it!is!not!known!whether!it!is!linear!or!not,!or!whether!the!sensor!behaves!similarly!in! different!temperatures.!Therefore!this!will!also!be!investigated!by!plotting!the!error!(Air!Quality!Egg!reading! minus!node!reading)!versus!temperature.!
4.3#Measurement#of#power#consumption#
Additionally! to! collecting! sensor! readings,! power! consumption! of! the! node! will! be! measured.! This! will! be! done!in!detail!to!see!what!additional!power!consumption!each!sensor!/!component!adds.!First,!the!power! consumption!of!the!basic!Seeeduino!Stalker!board!will!be!measured!with!no!peripherals!attached.!Further! on,! peripherals! will! either! be! attached! or! switched! on! in! code! one! by! one! and! the! change! in! power! consumption!noted.!Power!consumption!will!be!obtained!for!each!sensor!and!the!radio!and!the!Seeeduino! board.! It! will! be! measured! by! using! a! UNI/T! UT33D! multimeter! [32]! in! current! measuring! mode! and! connected! in! series! with! the! USB! power! supply! to! the! sensor! node.! The! current! consumption! can! be! multiplied!by!the!supply!voltage!(5!volts)!to!obtain!power!consumption!in!milliwatts.
5%–%Research%results%
!
Research! will! be! divided! into! two! main! parts:! Design! of! the! sensor! node! and! Measurement! results.! The! electronics!and!casing!design!will!be!discussed!first!and!measurements!and!performance!second.!
5.1 Design#of#the#Sensor#Node#
5.1.1%Circuit%design%
Two!versions!of!the!nodes!were!designed!circuit/wise,!one!with!the!option!to!switch!sensors!off!for!power! saving!and!the!second!using!a!simpler!circuit!but!without!this!option.!
The! power! saving! circuit! will! further! on! be! presented! in!Figure! 4! and! was! drawn! using! Scheme/it! online! software!by!the!electronics!supplier!Digi/Key![34].! ! ! ! Figure!4:!Circuit!schematic!of!the!power!saving!node!version! The!difference!between!the!power!saving!version!and!the!simpler!version!is!that!the!power!saving!version! uses! pin! D7! for! powering! the! nitrogen! dioxide! sensor! and! pin! D4! for! particle! sensor! readings! and! D5! for! powering! the! particle! sensor,! but! the! simpler! version! uses! pin! D7! for! particle! sensor! readings,! 5V! pin! for! powering!the!particle!sensor,!and!pin!D2!to!power!the!nitrogen!dioxide!sensor.!The!simpler!circuit!also!does! not!use!a!PNP!transistor!since!it!never!switches!off!the!particle!sensor.!Otherwise,!both!circuits!are!the!same! and!have!the!exact!same!functionality,!except!for!the!option!to!power!off!all!sensors!(except!the!CO!sensor)! on!the!board!in!Figure!4.!The!RES!(reset)!pin!of!the!board!is!connected!to!D11,!and!by!controlling!the!digital! pin,!it!is!possible!to!reset!the!board.!Since!the!node!only!has!to!send!sensor!readings!and!does!not!need!to! keep!track!of!absoulute!time,!a!reset!is!done!in!code!every!hour!as!maintenance.!Therefore!it!can!be!known! that!if!the!node!ever!“hangs”,!this!is!not!due!to!the!board!or!the!radio,!but!due!to!a!hardware!malfunction.!!A! picture!of!the!complete!electronics!connected!to!the!Seeeduino!board!is!given!in!APPENDIX!B.!!
5.1.2%Casing%and%installation%
!
The! casing! design! is! important! for! proper! installation,! and! emphasis! was! placed! on! weatherproofing! and! ease! of! maintenance! /! replacement.! The! circuit! was! placed! within! two! plastic! casings! for! extra! weatherproofing,!and!an!extra!box!was!placed!on!top!as!a!“roof”!to!protect!the!antenna!of!the!radio!from! rain.!As!an!example,!the!sensor!node!installed!at!Uppsala!Kommun!can!be!seen!in!Figure!5.!
The! blue! casing! is! the! inner! casing! and! the! transparent! one!is!the!outer.!The!inner!casing!houses!the!electronics! and! has! a! hole! in! the! bottom! for! air! to! pass! in! and! for! attaching!the!power!wire.!
The! outer! casing! is! glued! or! otherwise! fastened! to! the! wall! of! the! building,! and! the! inner! casing! placed! inside! and!held!by!the!antenna!screw!on!top.!The!outer!casing! has! a! large! opening! on! the! bottom! as! an! air! inlet.! The! inner! casing! is! attached! to! the! outer! casing! only! by! the! screw!of!the!antenna!on!the!top.!The!outer!casing!has!a! lid! facing! the! front! that! can! be! opened,! and! the! inner! casing! with! the! electronics! can! therefore! be! easily! swapped!by!simply!unscrewing!the!antenna.! ! ! Using!this!design!and!with!the!extra!box!on!top!for!rain!protection,!the!casing!is!weatherproofed!from!all! sides!except!from!the!bottom,!where!it!was!necessary!to!create!the!hole!for!air!inlet.!Both!gateways!that! receive!the!sensor!data!are!located!indoors!(one!in!the!Kommun!building!and!one!at!Stadsteater),!therefore! weatherproofing!for!these!is!not!required.!The!sensor!node!locations!are!marked!in!blue!in!Figure!6!and!the! location!of!the!existing!air!quality!measurement!unit!owned!by!the!Kommun!is!marked!in!yellow.! ! ! ! ! ! Figure!5:!Sensor!node!installed!at!Uppsala!Kommun!
5.2 Measurement#results#
Measurement! results! and! analysis! on! their! accuracy! will! be! presented.! The! sensors! that! are! factory! calibrated!will!not!be!analysed,!but!only!the!ones!requiring!calibration,!such!as!the!CO,!NO2!and!O3!sensors.!
Humidity! and! temperature! sensors! will! be! used! for! investigating! the! influence! of! these! parameters.! As! mentioned!earlier,!the!sampling!interval!for!all!sensors!is!approximately!3!minutes,!and!some!readings!may! be!missing!due!to!packet!loss,!which!will!not!be!discussed!here!but!instead!in!Aji!John!Pushpam’s!thesis.! !
5.2.1%Carbon%monoxide%sensor%readings%
The!sensor!used!is!pre/calibrated!to!an!extent,!but!may!need! scaling!for!sensitivity,!therefore!the!reading! without!calibration!is!at!worst offset by!±35%!as!given!by!the!manufacturer.!The!response!is!linear.
Carbon! monoxide! sensor! readings! were! observed! during! the! total! measurements! period! of! around! two! weeks.!All!of!the!sensors!gave!a!consistent!low!value:!zero!in!case!of!two!of!the!nodes!and!an!(uncalibrated)! average! value! of! around! 1! ppm! for! one! of! the! nodes.! The! sensor! nodes! had! been! verified! to! give! high! readings! by! using! smoke! of! burning! paper,! so! malfunction! was! excluded.! Also,! the! sensor! readings! of! the! reference!method!Air!Quality!Egg!were!consistent!with!these!results,!giving!a!constant!zero.!Since!the!Air! Quality!Egg!is!rated!to!sense!concentrations!above!0.5!ppm,!it!is!expected!that!the!carbon!monoxide!levels!in! the!city!were!below!this!detection!threshold.!Therefore!neither!the!Air!Quality!Egg!nor!the!nodes!developed! in! this! research! could! detect! such! low! concentration! levels.! Values! of! 1! ppm! and! below! are! also! safe! for! people! as! discussed! earlier,! therefore! these! results! confirm! good! air! quality! regarding! CO! concentrations.! Unfortunately!because!of!the!low!concentrations,!the!CO!sensors!of!the!node!of!this!research!could!not!be! field/calibrated! further! to! compensate! for! the! possible! ±35%! worst/case! offset.! This! will! be! further! on! discussed!in!future!work!and!calibration!would!require!a!known!gas!concentration!in!a!lab!environment.! The!observed!zero!CO!values!of!one!node!(at!Uppsala!Theater)!in!comparison!to!the!reference!Air!Quality! Egg!are!given!in!Figure!7!where!measurements!during!two!days!are!shown.!
! Figure!7:!CO!reading!comparison!between!node!at!theater!and!reference!Air!Quality!Egg
It!can!be!seen!that!the!Air!Quality!Egg!only!gives!higher!readings!during!stabilisation!after!being!switched!on.! Afterwards!readings!from!both!the!node!and!Air!Quality!Egg!are!sampled!frequently!and!are!always!zero.!
5.2.2%Nitrogen%dioxide%sensor%readings%
!
Calibration! on! the! data! of! the! nitrogen! dioxide! sensor! was! first! performed! as! described! in! the! previous! sections.! Afterwards,! data! of! two! days! (from! the! node! at! the! Uppsala! Theater)! was! compared! with! the! reference!Air!Quality!Egg.!The!comparison!is!shown!in!Figure!8.!
!
Figure!8:!NO2!reading!comparison!between!node!at!theater!and!reference!Air!Quality!Egg!
Several!results!can!be!inferred!from!this!data.!Firstly,!the!sensor!node’s!data!reaches!saturation!at!about!70! ppb!and!this!appears!to!be!the!maximum!measurement!range!for!this!particular!sensor!and!configuration.!As! discussed! before,! it! is! impossible! without! advanced! lab! equipment! to! estimate! the! sensitivity! of! a! sensor! beforehand,!and!there!is!large!manufacturing!variability!between!sensors.! Furthermore,!there!is!no!clear!obvious!correlation!between!the!data.!This!leads!to!the!theory!that!the!data! of!the!node!may!possibly!be!delayed!in!time!because!of!the!lack!of!an!air!fan,!as!discussed!in!the!Related! Work!section.!However,!the!time!shift!appears!to!be!several!hours,!which!is!impossible!given!the!small!air! volume,!so!the!possibility!of!a!time!shift!causing!the!data!to!not!correlate!is!discarded.!Even!though!the!data! does!not!obviously!correlate,!the!average!over!two!days!is!very!close:!57.05!ppb!for!the!node!and!46.84!ppb! for!the!Air!Quality!Egg.!Considering!that!the!units!are!parts!per!billion,!this!is!a!high!degree!of!accuracy!on! average.! The!correlation!coefficient!between!the!data!was!further!on!investigated.!There!was!almost!no!correlation,! with!R2=/0.17,!and!the!trends!were!even!opposite.!As!can!be!seen!in!Figure!8!readings!increase!for!the!node!
and!decrease!for!the!Air!Quality!Egg!and!vice!versa.!
The!error!for!the!NO2!readings!was!then!calculated!for!hourly!averages!to!estimate!whether!the!sensor!data!
! Figure!9:!Hourly!averages!of!NO2!readings!at!Uppsala!Theater! It!can!be!observed!that!the!largest!differences!occur!during!night/time.!The!largest!difference!is!37!ppb!at! 1:00!of!the!second!night.!It!is!likely!that!the!readings!are!influenced!by!temperature!and!/!or!humidity.! To!investigate!influence!of!temperature!and!humidity,!the!error!(Air!Quality!Egg!reading!minus!node!reading)! was!plotted!against!temperature!and!humidity!respectively.!This!was!done!to!see!how!error!correlates!with! changes! in! these! parameters.! Of! course,! since! field/calibration! is! used,! the! influence! of! other! parameters! such!as!changing!gas!concentration!cannot!be!excluded!from!these!results!(as!it!cannot!be!kept!constant).! However,!the!results!will!give!an!approximate!indication!of!how!temperature!and!humidity!correlate!with! measurement!error.!The!results!are!given!in!Figure!10!and!Figure!11.!
! Figure!10:!NO2!error!vs!temperature,!Uppsala!Theater!node
! Figure!11:!NO2!error!vs!humidity,!Uppsala!Theater!node
As!can!be!observed,!there!is!no!clear!patter!and!the!error!data!is!very!clustered.!This!indicates!that!at!the! same!temperature!/!humidity,!there!is!a!large!variation!in!data.!This!is!assumed!to!be!an!indication!that!the! data!of!this!particular!sensor!is!not!very!repeatable,!since!the!same!conditions!provide!different!results.!The! changing! gas! concentration! is! also! an! influential! factor,! however,! a! pattern! would! likely! be! visible! nevertheless.!
A!linear!fitting!was!performed!on!the!data!since!no!other!clear!trend!was!observable.!Then,!the!equation!of! the!linear!fit!(shown!as!the!y!equation!on!the!graphs)!was!used!to!perform!temperature!compensation!on! the! sensor! data.! This! compensation! is! of! course! approximate! because,! as! mentioned! before,! there! are! uncertain!parameters!in!field!calibration.!Also,!the!fit!of!the!linear!function!is!not!very!good,!as!indicated!by! the! R2! fit! values! of! less! that! 0.25! in! both! cases.! Nevertheless,! the! temperature! compensation,! even! if!
imperfect,! could! improve! data! accuracy,! and! therefore! was! applied.! Compensation! for! humidity! influence! was!excluded!from!the!scope!of!the!research,!as!it!was!found!to!be!a!less!influential!factor!in!most!cases.! The!temperature!compensated!data!is!shown!in!Figure!12.!
Even!though!the!fit!is!still!not!perfect,!and!the!correlation!was!calculated!to!be!very!low!(only!0.05),!the!data! still! follows! the! reference! much! closer! and! the! average! error! has! become! only! 0.05! ppb.! Therefore! for! a! day’s! measurements,! on! average! the! sensor! node! would! give! a! value! extremely! close! to! the! reference! method.!The!average!error!is!also!consistent!for!the!approx.!two!day!measurement!presented.!
It! was! seen! that! there! is! a! lot! of! variation! in! the! data,! therefore! it! was! investigated! whether! the! error! changes! with! temperature! by! calculating! the! average! error! in! each! temperature! range.! The! results! are! presented!in!Table!3.!
Table!3:!Temperature!and!average!error,!NO2!readings,!Uppsala!Theater!node!
Temperature,!°C!
Average!error,!ppb!
Average!error,!%!
7/8!
/37,86!
81,90!
8/9!
/31,75!
74,63!
9/10!
/16,70!
34,05!
10/11!
/14,14!
29,00!
11/12!
/4,28!
8,53!
12/13!
2,92!
6,60!
It!can!be!seen!that!the!error!drastically!increases!at!temperatures!below!11!degrees!Celsius,!and!that!the! error!is!relatively!low!in!higher!temperatures.! Further!on,!the!same!procedures!to!obtain!temperature!compensated!readings!were!performed!on!the!data! from!the!other!nodes!located!at!the!Uppsala!Kommun!(one!node!on!Kungsgatan!and!one!on!Vaksalagatan).! The!readings!are!shown!in!Figure!13!and!the!temperature!compensated!readings!are!shown!in!Figure!14.!In! the!temperature!compensated!reading!graph,!the!Air!Quality!Egg!readings!have!been!averaged!for!every!30! minutes!to!display!the!data!trends!more!clearly.! Figure!13:!NO2!concentration,!Kungsgatan!and!Vaksalagatan!nodes
! Figure!14:!Temperature!compensated!NO2!measurements,!Kungsgatan!and!Vaksalagatan!nodes
It!can!be!observed!that!before!temperature!compensation,!there!is!no!obvious!trend!/!correlation!between! the! two! node! and! reference! Air! Quality! Egg! measurements.! The! correlation! coefficient! between! the! Air! Quality! Egg! and! node! readings! was! calculated! to! be! /0.45! for! Kungsgatan! and! /0.75! for! the! Vaksalagatan! node.! A! negative! correlation! suggests! that! the! node! readings! decrease! when! the! Air! Quality! Egg! readings! increase,!and!there!isn’t!a!clear!or!consistent!correlation!because!both!nodes!have!a!different!correlation! coefficient.!
After!applying!temperature!compensation,!however,!the!Kungsgatan!node!has!0.95!correlation!with!the!Air! Quality! Egg! and! the! Vaksalagatan! node! has! 0.93.! The! data! now! correlates! really! well,! therefore! the! temperature!compensation!gave!a!huge!improvement!in!data!quality.!The!error!on!average!is!0.89!ppb!for! Kungsgatan! and! 4.18! ppb! for! Vaksalagatan! node.! The! average! error! is! small,! however,! it! can! be! observed! that!towards!the!end!of!the!presented!measurements,!the!error!becomes!larger!and!exceeds!20%!error!that! was!defined!as!an!accuracy!requirement.!Therefore!the!NO2!sensors!cannot!meet!the!accuracy!requirement,!
but! give! a! good! estimate! of! the! daily! average! value.! It! is! possible! that! the! cause! of! inaccuracy! is! manufacturing!variability!and!that!the!scaling!function!provided!by!the!manufacturer!is!not!suitable!for!every! sensor.!This!assumption!is!supported!by!the!fact!that!the!data!correlation!is!really!good,!therefore!the!scaling! of!the!data!could!be!altered!to!match!that!of!the!Air!Quality!Egg.!
It!also!has!to!be!noted!that,!when!comparing!the!data!from!the!Uppsala!Theater!node!and!the!other!two! nodes,!the!readings!of!the!other!two!nodes!were!taken!in!a!temperature!range!of!approx.!16/18!degrees! Celsius.! It! is! assumed! that! the! data! of! the! Uppsala! Theater! node! has! bad! correlation! because! the! lower! temperatures!(7/13!degrees)!caused!more!variation!in!the!data.!
5.2.3%Ozone%sensor%readings%
!
Calibration!of!the!ozone!sensor!was!similarly!performed!using!the!function!given!by!the!manufacturer,!as! described!in!the!previous!sections.!The!ozone!readings!from!a!week!of!measurements!of!the!sensor!node!at! Uppsala!Theater!and!Air!Quality!Egg!are!shown!in!Figure! 15.!The!few!missing!points!in!the!Air!Quality!Egg! graph!are!due!to!packet!loss.
! Figure!15:!O3!readings,!Uppsala!Theater!node
The! correlation! coefficient! between! the! data! was! calculated! to! be! /0.63.! Similar! to! the! readings! of! NO2!
presented! previously,! without! extra! compensation,! the! correlation! is! negative.! The! average! of! the! two! datasets! can! be! seen! to! be! close.! However,! the! sensor! node! at! first! appears! to! be! less! responsive,! which! causes!a!large!error,!at!certain!points!by!almost!100%.!
A!linear!temperature!compensation!equation!was!obtained!similarly!to!the!steps!done!for!the!NO2!sensors.!
The!data!after!temperature!compensation!is!presented!in!Figure!16.
It!is!apparent!that!the!temperature!compensation!provided!an!immense!improvement!in!data!accuracy.!The! correlation! coefficient! is! now! 0.81! and! the! average! error! is! 4.72%.! Except! for! a! couple! outlier! points,! the! maximum!error!never!exceeds!the!defined!requirement!of!20%!error.!
Furthermore,! an! interesting! phenomena! is! observed:! on! this! scale,! there! is! no! apparent! delay! or! lack! of! responsiveness!in!the!sensor!node’s!data.!The!sensitivity!of!the!node’s!sensor!is!also!good.!The!results!of!the! Air! Quality! Egg! and! the! node,! of! which! only! the! Egg! has! an! air! fan,! are! highly! comparable.! This! result! supports!the!assumption!that!the!use!of!a!fan!does!not!influence!sensor!readings!to!a!large!degree,!and!that! an!air!fan!is!not!mandatory!to!achieve!good!results.!
Results!from!a!two!day!measurement!from!the!node!at!Kungsgatan!are!presented!further!on.!Figure!17!is!the! data!before!and!Figure!18!is!the!data!after!temperature!compensation.!The!Air!Quality!Egg!data!in!the!second! figure!is!averaged!for!every!30!minutes!for!a!better!visibility!of!data!trends.!
! Figure!17:!O3!readings,!Kungsgatan!node
Again,! the! node! initially! appears! to! be! less! sensitive! to! ozone,! but! this! is! a! false! impression! since! after! temperature!compensation,!the!node!is!equally!sensitive!as!the!Air!Quality!Egg.!The!correlation!coefficient!is! similar! to! the! Uppsala! Theater! node:! 0.83,! and! the! average! error! is! only! 1.20%,! with! maximum! error! calculated!to!be!6.74%.!!
The!ozone!sensor!readings!have!therefore!been!proven!to!be!highly!comparable!to!the!reference!method! and!also!reliable!for!the!entire!measurement!time!(7!days!in!the!case!of!the!first!dataset).!The!importance!of! temperature!compensation!has!been!highlighted!for!both!the!ozone!and!nitrogen!dioxide!sensors.!Humidity! compensation! could! potentially! yield! even! more! accurate! results,! but! the! data! was! deemed! sufficiently! accurate,!and!humidity!compensation!will!rather!be!part!of!future!work.!Only!the!nitrogen!dioxide!sensor! showed!significant!variation!at!temperatures!below!10/11!degrees!Celsius,!so!this!is!an!issue!that!should!also! be!considered!in!the!future,!and!a!different!design!or!sensor!could!improve!accuracy.
5.2.4%Power%consumption
! ! Power!consumption!was!measured!using!the!UNI/T!UT33D!multimeter.!The!multimeter!was!set!in!current! measuring!mode!and!connected!in!series!to!the!5V!USB!power!supply!going!to!the!sensor!node,!as!explained! in!the!Research!Design!section.!The!current!draw!was!measured!with!±0.1mA!accuracy!(±0.1uA!accuracy!in! the!case!of!the!CO!sensor).!The! individual! power! consumption! of! the! various! components! was! calculated:! the! difference! was! taken! between! the! current! draw! of! the! system! before! and! after! turning! a! component! on.! The! steps! to! these! measurements!are!given!in!more!detail!in!APPENDIX!C.!Here,!just!the!results!are!presented!in!a!table!and! then!the!values!are!compared!to!what!is!given!in!the!datasheet!and!it!is!assessed!if!the!sensor!node!could!be! battery!powered.! ! Table!4:!Power!consumption!of!sensor!node!in!active!mode!
Component!/!Operation!mode!
Current!draw,!mA!
Only!the!Stalker!board!with!no!other!components! connected!and!running!the!program! 8.6! XBee!radio! 53.4! CO!sensor! 0.003! The!sensor!board!(all!sensors!disabled)! 14.8! Temperature!/!Humidity!sensor! 0.5! Ozone!sensor! 11.7! Nitrogen!dioxide!sensor! 9.4! Particle!concentration!sensor! 61.3! Total current!consumption:! 159.7! Table!5:!Power!consumption!of!sensor!node!in!sleep!mode!Component!/!Operation!mode!
Current!draw,!mA!
Only!the!Stalker!board!in!sleep!mode! 5.6! The!sensor!board!(all!sensors!disabled)! 14.8! Total!in!power!saving!mode:! 20.4! The!CO!board!was!measured!separately!to!consume!only!3uA,!which!is!likely!because!the!CO!sensor!actually! produces! current! instead! of! consuming! it! due! to! its! chemical! reaction! measuring! principle.! So! the! small! current!draw!is!actually!due!to!the!amplifier.!The!sensor!board!with!the!other!sensors!when!connected!but!disabled!draws!around!14.8mA.!!
