A Weather Ontology for Predictive Control in Smart Homes Paul - - PowerPoint PPT Presentation

▶
a weather ontology for predictive control in smart homes
SMART_READER_LITE
LIVE PREVIEW

A Weather Ontology for Predictive Control in Smart Homes Paul - - PowerPoint PPT Presentation

Oct 28, 2023 126 likes •465 views

Introduction Existing work Results A Weather Ontology for Predictive Control in Smart Homes Paul Staroch paulchen@rueckgr.at Arbeitsgruppe Automatisierungssysteme Institut fr Rechnergesttzte Automation Supervisors: Ao.Univ.-Prof.

slide-1
SLIDE 1

Introduction Existing work Results

A Weather Ontology for Predictive Control in Smart Homes

Paul Staroch

paulchen@rueckgr.at Arbeitsgruppe Automatisierungssysteme Institut für Rechnergestützte Automation Supervisors: Ao.Univ.-Prof. Dipl.-Ing. Dr.techn. Wolfgang Kastner Dipl.-Ing. Mario Kofler

October 10, 2013

slide-2
SLIDE 2

Introduction Existing work Results

Outline

Introduction Existing work Ontologies Weather data Ontology design methodologies Results SmartHomeWeather Weather Importer Conclusion

slide-3
SLIDE 3

Introduction Existing work Results

Smart Homes

  • Smart homes are equipped with some kind of intelligence

to perform tasks on their own.

  • Components: Sensors, actuators, communications

network, intelligent control. Goals:

  • Support with routine tasks.
  • Maintaining or increasing comfort.
  • Reduction of energy consumption.
slide-4
SLIDE 4

Introduction Existing work Results

Problems of smart homes

There are many smart home projects: Mozer’s adaptive house, Georgia Tech Aware Home, Gator Tech Smart Home, . . . However, in many cases there are several problems:

  • High complexity.
  • Optimisations and customisations are difficult.
  • Missing powerfulness and flexibility.

In many cases, the full potential of smart homes is not exploited.

slide-5
SLIDE 5

Introduction Existing work Results

An ontological approach

slide-6
SLIDE 6

Introduction Existing work Results

Weather data

Processes in and around a dwelling influenced by weather, e.g.:

  • Heating, ventilation, and air conditioning (HVAC).
  • Optimal utilisation of solar and wind power.
  • Irrigation.
  • Preparations for severe weather.

SmartHomeWeather is an ontology covering current and future weather data.

slide-7
SLIDE 7

Introduction Existing work Results

Outline

Introduction Existing work Ontologies Weather data Ontology design methodologies Results SmartHomeWeather Weather Importer Conclusion

slide-8
SLIDE 8

Introduction Existing work Results

Weather ontologies

Several ontologies cover weather data:

  • Semantic Sensor Web
  • SSN Ontology
  • SWEET
  • NNEW
  • . . .

Unfortunately, none of them was found to be suitable for smart homes.

slide-9
SLIDE 9

Introduction Existing work Results

Related ontologies

  • Location: Basic WGS84 (lat/lon) Vocabulary
  • Date and time: OWL-Time
  • Units of Measurement:
  • Measurement Units Ontology
  • Ontology of Units of Measure and Related Concepts
  • . . .

However, all these ontologies come with various drawbacks.

slide-10
SLIDE 10

Introduction Existing work Results

Measurement Units Ontology (1)

Weather phenomenon temperature 17.2^^xsd:float hasT emperatureValue rdf:type

slide-11
SLIDE 11

Introduction Existing work Results

Measurement Units Ontology (2)

Weather phenomenon temperature 17.2^^xsd:float hasT emperatureValue rdf:type muo:numerical value muo:Quality value rdf:subPropertyOf muo:Quality value rdfs:subClassOf muo:degree Celsius muo:measured in muo:Unit of measurement rdf:type

slide-12
SLIDE 12

Introduction Existing work Results

Outline

Introduction Existing work Ontologies Weather data Ontology design methodologies Results SmartHomeWeather Weather Importer Conclusion

slide-13
SLIDE 13

Introduction Existing work Results

Sensors and services

SmartHomeWeather retrieves data from local weather sensors and Internet weather services.

  • Arbitrary number of sources possible.
  • Assignment of priority values to weather data.
  • Current data from sensors and services.
  • Forecast data from services.
  • Time range for forecasts: 24 hours.
slide-14
SLIDE 14

Introduction Existing work Results

Weather sensors

Sensors are commonly accessed via fieldbus systems (KNX, LonWorks, BACnet, . . . ). A variety of sensors is available:

  • Barometer
  • Photometer
  • Hygrometer
  • Rain gauge
  • Pyranometer
  • Thermometer
  • Wind wane, anemometer
slide-15
SLIDE 15

Introduction Existing work Results

Weather services

  • Weather services evaluated: DWD, Google Weather Feed,

METAR, NWS, Weather.com, Weather Underground, World Weather Online, Yahoo! Weather, yr.no.

  • Criteria for evaluation: Coverage area, data format, data

access, access restrictions, terms of use, documentation, stability, weather elements, time frame, weather updates.

  • Conclusion: Reference implementation using yr.no
slide-16
SLIDE 16

Introduction Existing work Results

Weather elements

Weather elements currently used in SmartHomeWeather:

  • Temperature
  • Relative humidity
  • Dew point
  • Cloud coverage (altitude and amount cloud cover)
  • Precipitation (intensity and probability)
  • Wind (speed and direction)
  • Atmospheric pressure
  • Solar radiation
  • Position of the sun (azimuth, elevation angle)
  • Weather condition (sunshine, rain, snow, . . . )
slide-17
SLIDE 17

Introduction Existing work Results

Outline

Introduction Existing work Ontologies Weather data Ontology design methodologies Results SmartHomeWeather Weather Importer Conclusion

slide-18
SLIDE 18

Introduction Existing work Results

Methodologies

  • Ontology 101
  • Uschold and King
  • TOronto Visual Enterprise
  • UPON
  • METHONTOLOGY
slide-19
SLIDE 19

Introduction Existing work Results

METHONTOLOGY

States Conceptualisation Planification Activity Knowledge Acquisition Documentation Evaluation Activities Formalisation Integration Specification Implementation Maintenance

slide-20
SLIDE 20

Introduction Existing work Results

Outline

Introduction Existing work Ontologies Weather data Ontology design methodologies Results SmartHomeWeather Weather Importer Conclusion

slide-21
SLIDE 21

Introduction Existing work Results

Competency questions

  • What will the weather situation be in one hour, in two

hours, . . . , in 24 hours?

  • What will be the minimum temperature, humidity, . . . over

the next 24 hours? What about maximum values?

  • Will the weather change? Will the temperature, humidity,

. . . rise or fall?

  • Does it rain? Will it rain in the next hours? Will it rain

today?

  • Will temperature drop/stay below 0 ◦C?
  • When can we open windows and when do we have to keep

them shut?

slide-22
SLIDE 22

Introduction Existing work Results

Overview

Weather report Weather report source Weather state Weather phenomenon Weather condition is source of has source has condition has weather state belongs to report has weather phenomenon belongs to state has previous weather state has next weather state

slide-23
SLIDE 23

Introduction Existing work Results

Concept hierarchies: Weather phenomenon

Weather phenomenon rdfs:subClassOf Atmospheric pressure Cloud cover Precipitation Temperature Wind rdfs:subClassOf rdfs:subClassOf rdfs:subClassOf rdfs:subClassOf Dew point rdfs:subClassOf Humidity rdfs:subClassOf Sun position rdfs:subClassOf Solar radiation rdfs:subClassOf

slide-24
SLIDE 24

Introduction Existing work Results

Concept hierarchies: Temperature

Temperature rdfs:subClassOf Below room temperature Heat Above room temperature Frost rdfs:subClassOf rdfs:subClassOf rdfs:subClassOf Cold rdfs:subClassOf Room temperature rdfs:subClassOf

slide-25
SLIDE 25

Introduction Existing work Results

Concept hierarchies: Weather report

Weather report rdfs:subClassOf Weather report from service Weather report from sensor rdfs:subClassOf Forecast weather report Current weather report from sensor Short range forecast rdfs:subClassOf rdfs:subClassOf Current weather report rdfs:subClassOf rdfs:subClassOf rdfs:subClassOf Current weather report from service Medium range forecast Long range forecast rdfs:subClassOf rdfs:subClassOf rdfs:subClassOf rdfs:subClassOf

slide-26
SLIDE 26

Introduction Existing work Results

Concept hierarchies: Weather state

Weather state Calm weather Clear weather Cloudy weather Cold weather Dry weather Hot weather Moist weather No rain weather Pleasant temperature weather Rainy weather Windy weather Sun protection weather Fair weather Airing weather Very rainy weather No awning weather Severe weather Stormy weather Thunderstorm

slide-27
SLIDE 27

Introduction Existing work Results

SPARQL and SWRL (1)

SELECT ?s WHERE { ?s weather:hasWeatherPhenomenon ?p. ?p a weather:Frost. ?s weather:belongsToWeatherReport ?r. ?r a weather:ShortRangeForecastReport. }

slide-28
SLIDE 28

Introduction Existing work Results

SPARQL and SWRL (2)

hasWeatherPhenomenon(?s1, ?t1) ∧ hasTemperatureValue(?t1, ?v1) ∧ numericalValue(?v1, ?m1) ∧ hasWeatherPhenomenon(?s2, ?t2) ∧ hasTemperatureValue(?t2, ?v2) ∧ numericalValue(?v2, ?m2) ∧ greaterThan(?m2, ?m1) ∧ hasNextWeatherState(?s1, ?s2)

⇒ increasingTemperature(?s1, ?s2)

slide-29
SLIDE 29

Introduction Existing work Results

Outline

Introduction Existing work Ontologies Weather data Ontology design methodologies Results SmartHomeWeather Weather Importer Conclusion

slide-30
SLIDE 30

Introduction Existing work Results

Weather Importer

WeatherPhenomenon WeatherReport

  • priority: int

WeatherState Weather

  • priority: int

1..1 0..* 1..1 1..1 0..1 0..1 0..1 0..1 1..1 0..*

  • Import from sensors and Internet services.
  • Unit tests for SmartHomeWeather and Weather Importer.
slide-31
SLIDE 31

Introduction Existing work Results

Outline

Introduction Existing work Ontologies Weather data Ontology design methodologies Results SmartHomeWeather Weather Importer Conclusion

slide-32
SLIDE 32

Introduction Existing work Results

Conclusion

Results:

  • SmartHomeWeather allows predictive control based on

weather data within smart homes.

  • Weather Importer retrieves weather data from various

sources into SmartHomeWeather. Future work:

  • Interoperability with other data sources.
  • Smart Cities.
slide-33
SLIDE 33

Introduction Existing work Results

The End

Thanks for your attention. Questions?