Semantic Data Layers in Air Quality Monitoring for Smarter Cities

Semantic Data Layers in Air Quality Monitoring for Smarter Cities

Jean-Paul Calbimonte, Julien Eberle and Karl Aberer

LSIR EPFL

S4SC Workshop 2015. International Semantic Web Conference ISWC

Bethlehem, PA, October 2015

@jpcik

5% (2.6 million) of all deaths are caused by urban air pollution

Global Health Risks, WHO 2014

Health studies show that air pollution increases the risk of cardiovascular mortality(heart attacks) by 5% to 20% at least

Health studies have shown the link between pollution and cardiovascular mortality

Cardiopulmonary Diseases

CardiovascularDiseases

Brook et al, Circulation 2010

Ischemic HeartDiseases

cardiovascular & respiratory morbidity

negative effects on nervous system

COcarbon monoxide

respiratory morbidity

airway hyper responsiveness

harmful to living organisms

decreased lung function

lung inflammationUrbain Air Pollutants

NOxnitrogendioxide

monoxide

O3 ozonePMparticulatematter

aggravation pulmonary &

cardiovascular condition

Air Pollution in a Smart City

• Collaborative acquisition• Self-diagnosis• Localized measurements• Heterogeneous data sources• City Health Studies• Open Air Quality Index• Citizen Science• Data Privacy

OpenSense IISensing the air we breath

http://opensense.epfl.ch/

OpenSense2global concernhighly location-dependenttime-dependent

Crowdsourcing High-Resolution Air Quality Sensing

Air Pollution

Accurate location-dependent and real-time information on air pollution is needed

Integrated air quality measurement platform

Heterogeneous devices and data Human activity assessment, lifestyle and health data

• Link high-quality and low-quality data • Integration of pure statistical models and physical

dispersion models• Better coverage through crowdsensing• Incentives for crowd data provision• Finer temporal and spatial resolutions• Utilitarian approach for trade-off between model

complexity, privacy and accuracy• Higher accuracy of pollution maps models

http://opensense.epfl.ch

Institutional stations

OpenSense infrastructure

Personal mobile sensors CrowdSense

2 km

Governmental station

Our measurements

Zürich: 10 sensor nodes updated: O3, NO2, CO, UFP, GSM, GPS

EMPA/Decentlab: 2 sensor nodes (Aircubes) : 2x O3, 3x NO2EMPA/Decentlab: comparing Aircubes with reference

Lausanne Deployment

Particle sampling module• Ultrafine particle

measurements using NaneosPartector

• Measures directly lung-deposited surface area

Gas sampling module• CO, NO2, O3, CO2, temperature

& relative humidity • Hybrid active sniffer/closed

chamber sampling operation• Enables absolute

concentration mobile measurements

Enhanced localization & logger• mounted inside bus• Fused GPS, gyro and vehicle

speedpulses• Accurate sample geolocation even in

difficult urban landscapes• GPRS communication

Beyond Public Transportation: C0

Crowdsensing• Participatory Sensing• Data Aggregation• Data privacy / utility• Combine with sensing

infrastructure

Use smartphones to automatically

gather the contextual information

- location, activity and

environment

- Data is aggregated

anonymously and fused with other sources

Crowdsensing

Reference station

Crowd sensing

Public transportation

Raw Data Acquisition

Air Pollutants Time Series

Temporal Spatial Aggregations

Pollution Maps Pollution ModelsAir Quality

recommendationsHealth Studies

Air Quality Products &

Applications

From Sensing to Actionable Data

Localization: GNSS fusioned with odometry

GPRS

• packet parser• system logging• database server• GPS interpolation• advanced filtering• fault detection• system health monitor• automatic reporting

10

bu

ses

in L

ausa

nn

eCO, NO2, O3, CO2, UFP, temperature, humidity

The Lausanne Deployment

GSN: Global Sensor Networks

https://github.com/LSIR/gsn

GSN: Accessing the Data

18

• Public Data Access• Online Processing• Basic Plotting/Filtering• Further Processing

Data Formats: CSV on the Web

19

{"name": "type_event","virtual": true,"aboutUrl": "#obs-{_row}","propertyUrl": "rdf:type","valueUrl": "opensense:CO_Observation"

}, {"name": "unit","virtual": true,"aboutUrl": "#obs-{_row}","propertyUrl": "qu:unit","valueUrl": "unit:mgm3"

}]}

}

{"@context": ["http://www.w3.org/ns/csvw", {"@language": "en"}],"url": "opensense.csv","tableSchema": {

"columns": [{"name": "time","titles": "Time","aboutUrl": "#obs-{_row}","propertyUrl": "ssn:observationResultTime"

"datatype": {"base": "datetime","format": "yyyy-MM-ddTHH:mm" },

}, {"name": "station","titles": "Bus sensor","aboutUrl": "#obs-{_row}","propertyUrl": "ssn:observedBy"

}, {"name": "co","titles": "CO concentration","aboutUrl": "#obs-{_row}","propertyUrl": "ssn:observationResult"

}http://www.w3.org/2013/csvw/

Example: SSN Ontology

20

SSN Ontology

21

ssn:Sensor

ssn:Platform

ssn:FeatureOfInterest

ssn:Deployment

ssn:Property

cf-prop:air_temperature

ssn:observes

ssn:onPlatform

dul:Place

dul:hasLocation

ssn:SensingDevicessn:inDeployment

ssn:MeasurementCapability

ssn:MeasurementProperty

geo:lat, geo:lng

xsd:double

ssn:hasMeasurementProperty

ssn:Accuracy

ssn:ofFeature

aws:TemperatureSensor

aws:Thermistor

ssn:Latency

dim:Temperature

qu:QuantityKind

cf-prop:soil_temperature

cf-feat:Wind

cf-feat:Surface

cf-feat:Medium

cf-feat:aircf-feat:soil

dim:VelocityOrSpeedcf-prop:wind_speedcf-prop:rainfall_rate

aws:CapacitiveBead …

…

…

t1t2

t3t4

aggregated

Pro

ven

ance

Raw data

Spatio-temporal aggregation

Event Annotation

Temporal Segmentation

Spatial Processing

Data Granularity & Annotation

From Activity Recognition To Exposure

• Accelerometer / Location Data• Online Activity Detection• Online Breathing / intake estimation• Estimation of Air Pollutant Exposure• Personalized recommendations

TinyGSN

24

• Android background application

• Front-end to change parameters

and select the sensors to use.

• Based on the same principles as

GSN:

• wrapper,

• virtual sensors

• streamElements.

• Scheduler, optimized for gathering

continuous location without

depleting the battery,

• Manage Android Services and

Alarms

• Allow the device to sleep between the measurements.

Goal: estimate the health effects of long-term exposure to air pollution

Health Studies in OpenSense

Health Studies in OpenSense

Combine with Pollution Models

Data filtering & calibration

Data validation

LUR model

Pollutionmap

Raw data

Processing steps:

Mapvalidation

2 km

Governmental station

Our measurements

From raw measurements to fine-grained pollution maps

Winter (January - March)Spring (April – June)Summer (July - September)Fall (October – December)

OpenSense: Air Pollution in Smart cities

• Data quality and curation• Failure and Noise handling• Semantic Layers of Data for different purposes• Privacy protection• Air qualitymodels and data fusion• Incentives for participatory sensing• Personalized health recommendations

Muchas gracias!

Jean-Paul CalbimonteLSIR EPFL

@jpcik