Muhammad Rizwan Saeed, Charalampos Chelmis, Viktor K. …€¦ · Muhammad Rizwan Saeed,...

1
Contact information: USC Data Science Lab http://dslab.usc.edu/ Center for Interactive Smart Oilfield Technologies http://cisoft.usc.edu/ SOSNet: Smart Oil Field Safety Net Semantic Data Management for the Smart Oil Fields Muhammad Rizwan Saeed, Charalampos Chelmis, Viktor K. Prasanna An integrated system that automatically: • integrates multiple heterogeneous streams of data • performs analytics for early identification of Loss of Containment (LOC) events • facilitates decision making for asset integrity experts Sensory Data Decision Makers SOSNet Solution Structured, unstructured sources Integrated Asset Integrity Management (IAIM) Data Integration Ontology-guided automatic integration • Create semantic models (ontologies) of the domain of interest to annotate raw data sources • Use ontologies for automatic integration of heterogeneous structured and semi-structured data sources reads raises Minimum Thickness CorrosionDetectedAlert hasCharacteristics RateCalculations raises Visual Analytics reads Predictive Analytics raises measuredAs hasMeasurement Equipment Section Measuring Location hasPart hasPart EquipmentCategory iso:vessel iso:tank hasGroup sofos:Facility hasEquipment Photographs Visuals measuredAs Findings CorrosionLevels ouputs Manual Analysis ThicknessTS Is part of TextTS ImagesTS Is part of Is part of reads Categories raises LOCPredictedAlert raises CriticalWO generates generates Welding Painting Repair Corrosion Thickness ThicknessValue hasMeasurement ouputs hasResolution iso:pipe Iso:physicalObject rdfs:rdfs:subClassOf sofos:ObservedEntity rdfs:subClassOf poem:Measurement rdfs:subClassOf poem:Observation rdfs:subClassOf reads reads sofos:analyticsModule rdf:typeOf sofos:analyticsModule rdf:typeOf rdf:typeOf schema:imageObject schema:Text subClassOf subClassOf Smart Oilfield Ontology Data Integration, Querying, and Analytics • Deal with data heterogeneity • Build smart browse and search capabilities • Abstract users from underlying standards and languages • Provide domain experts with an optimized and integrated view of the environment Data Querying ASQFor: Automatic SPARQL Query Formulation • Allows user to intuitively query a semantic database • Lets user formulate formal SPARQL queries indirectly • User performs intuitive keywords search • System translates it to SPARQL query Data Querying ASQFor: Automatic SPARQL Query Formulation • Benefits: • No knowledge of schema and query language required for end user. • Domain independent portable database search application Query formulation time is fairly consistent irrespective of the size of the repository Data Analytics Similarity-based queries (Current research) • Content-based recommender system for the smart oilfield SELECT DISTINCT ?facilityname ?equipmentname ?tml_txt ?minthick ?measurementdate ?thickness ?user from <http://localhost:8890/SOSNETDEMO0411162300> WHERE { ?facility <http://www.usc.edu/cisoft/sosnet/sofos#hasEquipment> ?equipment. ?equipment <http://www.usc.edu/cisoft/sosnet/sofos#hasTML> ?tml. ?tml <http://www.usc.edu/cisoft/sosnet/sofos#hasAttributes> ?attributes. ?tml <http://www.usc.edu/cisoft/sosnet/sofos#hasMeasurements> ?measurements. ?measurements <http://www.usc.edu/cisoft/sosnet/sofos#hasObservation> ?observation. ?facility <http://www.usc.edu/cisoft/sosnet/sofos#FacilityName> ?facilityname. ?equipment <http://www.usc.edu/cisoft/sosnet/sofos#EquipmentName> ?equipmentname. ?tml <http://www.usc.edu/cisoft/sosnet/sofos#TML_TXT> ?tml_txt. ?attributes <http://www.usc.edu/cisoft/sosnet/sofos#MINTHICK> ?minthick. ?observation <http://www.usc.edu/cisoft/sosnet/sofos#MeasurementDate> ?measurementdate. ?observation <http://www.usc.edu/cisoft/sosnet/sofos#Thickness> ?thickness. ?observation <http://www.usc.edu/cisoft/sosnet/sofos#User> ?user. FILTER ( ?facilityname = "EI252I" ) } Comparison between query formulation and query execution times using ASQFor Search Key Results KBC-0101 KBC-0103 KBC-0108-CP KBC-0103-CP KTT-0109 KAN-0106 Same drawing, equipment type, location, and corrosion level VESSEL ABH-100 type has-survey Survey0100 3/3/2003 has-date NORMAL status WO230010 has-WO ET-200 has-equipment OIL fluid Found. has-finding "Perform..." "Painted..." problem recommendation solution Found. description VESSEL MBH-200 type has-survey Survey0220 3/3/2001 5/2/2005 NORMAL status WO231040 has-WO KP-100 has-equipment OIL fluid "Corroded.." has-finding "Perform..." "Replaced.." problem recommendation solution "Fix..." description similar-to

Transcript of Muhammad Rizwan Saeed, Charalampos Chelmis, Viktor K. …€¦ · Muhammad Rizwan Saeed,...

Page 1: Muhammad Rizwan Saeed, Charalampos Chelmis, Viktor K. …€¦ · Muhammad Rizwan Saeed, Charalampos Chelmis, Viktor K. Prasanna An integrated system that automatically: •integrates

Contact information:USC Data Science Lab http://dslab.usc.edu/Center for Interactive Smart Oilfield Technologies http://cisoft.usc.edu/

SOSNet: Smart Oil Field Safety Net

Semantic Data Management for the Smart Oil FieldsMuhammad Rizwan Saeed, Charalampos Chelmis, Viktor K. Prasanna

An integrated system that automatically:

• integrates multiple heterogeneous streams of data

• performs analytics for early identification of Loss of Containment (LOC) events

• facilitates decision making for asset integrity experts

Sensory Data Decision Makers

SOSNet Solution

Structured,

unstructured

sources

Integrated Asset Integrity Management (IAIM)

Data Integration

Ontology-guided automatic integration

• Create semantic models (ontologies) of the domain of interest to annotate raw data sources

• Use ontologies for automatic integration of heterogeneous structured and semi-structured data sources

reads

raises

Minimum Thickness

CorrosionDetectedAlert

hasCharacteristics

RateCalculations

raises

Visual Analytics

reads

Predictive Analytics

raises

measuredAs

hasMeasurement

Equipment

Section

Measuring Location

hasPart

hasPart

EquipmentCategory

iso:vessel

iso:tankhasGroup

sofos:Facility

hasEquipment

Photographs

VisualsmeasuredAs

Findings

CorrosionLevels

ouputs

Manual Analysis

ThicknessTS

Is part of

TextTS ImagesTS

Is part of Is part of

reads

Categoriesraises

LOCPredictedAlert

raises

CriticalWO

generates

generates

WeldingPainting Repair

Corrosion

Thickness

ThicknessValue

hasMeasurement

ouputs

hasResolution

iso:pipe

Iso:physicalObject

rdfs:rdfs:subClassOf

sofos:ObservedEntity

rdfs:subClassOf

poem:Measurement

rdfs:subClassOf

poem:Observationrdfs:subClassOf

reads

readssofos:analyticsModulerdf:typeOf

sofos:analyticsModule rdf:typeOf

rdf:typeOf

schema:imageObject

schema:TextsubClassOf

subClassOf

Smart Oilfield Ontology

Data Integration, Querying, and Analytics

• Deal with data heterogeneity

• Build smart browse and search capabilities

• Abstract users from underlying standards and languages

• Provide domain experts with an optimized and integrated view of the environment

Data Querying

ASQFor: Automatic SPARQL Query Formulation

• Allows user to intuitively query a semantic database

• Lets user formulate formal SPARQL queries indirectly

• User performs intuitive keywords search

• System translates it to SPARQL query

Data Querying

ASQFor: Automatic SPARQL Query Formulation

• Benefits:

• No knowledge of schema and query language required for end user.

• Domain independent portable database search application

• Query formulation time is fairly consistent irrespective of the size of the repository

Data Analytics

Similarity-based queries (Current research)

• Content-based recommender system for the smart oilfield

SELECT DISTINCT ?facilityname ?equipmentname ?tml_txt ?minthick?measurementdate ?thickness ?user from <http://localhost:8890/SOSNETDEMO0411162300> WHERE { ?facility <http://www.usc.edu/cisoft/sosnet/sofos#hasEquipment> ?equipment. ?equipment <http://www.usc.edu/cisoft/sosnet/sofos#hasTML> ?tml. ?tml <http://www.usc.edu/cisoft/sosnet/sofos#hasAttributes> ?attributes. ?tml <http://www.usc.edu/cisoft/sosnet/sofos#hasMeasurements> ?measurements. ?measurements <http://www.usc.edu/cisoft/sosnet/sofos#hasObservation> ?observation. ?facility <http://www.usc.edu/cisoft/sosnet/sofos#FacilityName> ?facilityname. ?equipment <http://www.usc.edu/cisoft/sosnet/sofos#EquipmentName> ?equipmentname. ?tml <http://www.usc.edu/cisoft/sosnet/sofos#TML_TXT> ?tml_txt. ?attributes <http://www.usc.edu/cisoft/sosnet/sofos#MINTHICK> ?minthick. ?observation <http://www.usc.edu/cisoft/sosnet/sofos#MeasurementDate> ?measurementdate. ?observation <http://www.usc.edu/cisoft/sosnet/sofos#Thickness> ?thickness. ?observation <http://www.usc.edu/cisoft/sosnet/sofos#User> ?user. FILTER ( ?facilityname = "EI252I" ) }

Comparison between query formulation and query execution times using ASQFor

Search Key Results

KBC-0101 KBC-0103KBC-0108-CPKBC-0103-CPKTT-0109KAN-0106

Same drawing, equipment type, location, and

corrosion level

VESSEL

ABH-100

type

has-survey

Survey0100

3/3/2003

has-date

NORMAL

status

WO230010

has-WO

ET-200

has-equipment

OIL

fluid

Found .

has-finding

"Perform..."

"Painted..."

problem recommendation

solution

Found .

description

VESSEL

MBH-200

type

has-survey

Survey0220

3/3/2001

5/2/2005

NORMAL

status

WO231040

has-WO

KP-100

has-equipment

OIL

fluid

"Corroded.."

has-finding

"Perform..."

"Replaced.."

problem recommendation

solution

"Fix..."

description

similar-to