performing event detection over real-time sensor data using ontology-driven approaches csiro land...
TRANSCRIPT
Performing event detection over real-time sensor data using ontology-driven approaches
CSIRO LAND AND WATER
Jonathan Yu | Research software engineerEnvironmental Information Systems, CLW Highett
14 May 2013
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Outline
Background: Event detection over real-time sensor data
Capturing machine readable semantics – ontologies
Enabling user-based events detection using ontology-driven approaches
2 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Performing event detection over real-time sensor data
Real-time observations are increasingly becoming available through sensor networks• Reporting, monitoring, analysis
Examine events that happen in a sensor network and get notifications
• Mitigate risk in the environment
• Improve management response times
3 |
WQ Weather Flow
Sensor Network
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Water quality issues...
Example events in this domain:• Total nitrogen conc. in a river > X mg/L• Dissolved oxygen conc. at sensor < Y mg/L
4 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Undetected sewer rising mains pipe failures...
Direct costs: water service providers ($ mil. per event)
Indirect costs: social, environmental ($10k - $1 mil. per event)
5 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Sewer rising mains case study
6 |
0 2000 4000 6000 8000 10000 12000 14000 160000
20
40
60
80
100
120
140
160
Time (mins)
Flow
rate
(l/s
)
Pipe failure event = flow > 100 l/s
Example event: Flow rate > 100 l/s
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Sensor middleware - Global Sensor Network (GSN)
WQ WeatherFlow
Sensor Network
GSN
Virtual Sensor (WQ)
Virtual Sensor(Flow)
Virtual Sensor (Aggr.)
End users
User Interface
7 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Stepped notifications: < 5 consecutive = low, 5 - 20 consecutive = moderate, > 20 = high risk
Using GSN for event detection on sewer rising mains
(Low risk)
(High risk)
Stepped Notifications
Flow observations
Simple Moving Average
Looking for when flow exceeds a preset threshold over the Simple Moving average
8 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Existing workflow: real-time event detection
High level entry for an end user e.g. Scientists and managers• Inefficient
Knowledge hidden behind code or in people’s heads - implicit semantics• Barrier for reusability• Possible inconsistencies
9 |
Curation CodingAnalysis,
Monitoring, Management
SensorMiddleware
(GSN)
Sensor Network
End users
Programmers
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Problem of data heterogeneity, integration
Multiple datasets• Often multiple data schemas, formats, field names,
conventions
The use of the observation property “Flow rate”• Flow• FLOW_RATE• RATE_OF_FLOW_L_per_s
Need a mechanism for consistent use of semantics• Map to shared and commonly understood definitions
10 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Capturing semantics, not just syntax
Syntax of a language refers to the structure or grammarSemantics of a language refers to the meaning
E.g. “Colorless green ideas sleep furiously.”
• Syntactically correct• Semantically meaningless/inconsistent
11 |
Zzzzzz...
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Enable capture and consistent use of semantics
12 |
Observation event at Flow Sensor X
Flow
PVC Pipe at Clunies Ross Street
100 litres per second
Flow sensor X
Flow rate sensing
Has an observation result
Some result
Has value
Produced by
implementsobserves
Has observation property
Has feature of interest
Observation event at WQ sensor
Lake Burley Griffin
Another result 10mg/L
WQ meter
Dissolved oxygen sensingDissolved
oxygen conc.
Observation
Feature of interest
Sensor Output Observation Value
Sensor
SensingProperty
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
OntologiesEnable specification of semanticse.g. Concepts, relationships, logic
assertions
Provides ability to refer to ‘Flow rate’ concept (semantics), rather than FLOW_RATE (syntax)
Machine readable/processable• Using Web Ontology Language (OWL)• W3C standard – “semantic web”
13 |
implementsobserves
Sensor
SensingProperty
Observation
Sensor Output
Has an observation
result
Produced by
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Leverage ontology tools
• Query languages like “SPARQL” – SPARQL Protocol and RDF Query
Language– For querying RDF statements
which OWL builds upon
– E.g. Find all sensors that observe flow rate
• Triple stores for storing RDF and ontology statements
• Ontology reasoners– Automated consistency checking– Inference engines
All men are mortal,Socrates is a man=> Socrates is mortal
All sewer pipe bursts near rivers cause some environmental damage,
Pipe X is near River Y,Pipe X has a pipe burst event=> River X has caused some
environmental damage at River Y
14 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Enabling “user-based” real-time event detection
Enable capture of user-defined semantics of events & reference sensors consistently
1) Sensor network system semantics described (e.g. Flow rate sensor is located at X)
2) Domain of interest semantics described (e.g. Flow is an observable property)
3) Event semantics described (e.g. Flow rate at sensor#1 > 100 l/s)
4) Machine-readability: for rendering in user interfaces & code generation
15 |
SensorMiddleware
(GSN)
Sensor Network
End users
?
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Return notifications from triggered events with metadata based on ontology semantics
e.g. The sewer rising main has a problem due to <...>
Ontology-driven event detection system
16 |
SensorMiddleware
(GSN)
Sensor Network
End users
Ontology-enabledUser Interface
OntologiesSemantic
Sensor Net. Ontology
DomainOntology
Annotates available sensors and their capabilities
e.g. Flow rate sensor data at Location X
Generate code for event detection using event constraint semanticse.g. FLOW_RATE > 100 l/s
Populate user interface elements based on domain semantics and sensor network annotations.
Allow users to define event constraints
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Semantic Sensor
Network
Event-detection
WQdomain
Urban water domain
WQ user Urban water user
Mid-Upper ontologies
Application ontologies Domain ontologies
User ontologies
Representing domains and applications
???
17 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Event Detection Ontology def’s:Event Rule, Value Constraints, Units
18 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Domain ontologies (uwda:) - Sensors
19 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Event rule definition instances
Rule ID Observed property
Value constraint
Feature of interest
Observed By (Sensor)
1 Flow
2 Flow > 100 l/s
3 Flow > 100 l/s Pipe A
4 Flow > 100 l/s Pipe Sensor A-1
5 Flow > 100 l/s Pipe A Pipe Sensor A-1
20 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Event dashboard - User interface demo for urban water domain
21 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu22 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Ontology APIs
Ontology-driven user interface
23 |
SensorMiddleware
(GSN)
Ontology-enabledUser Interface
Triple store(User ontologies)
Ontology Reasoner
Ontology definitions
End usersQuery/Rule engines
Presentation Widgets(Standard web forms)
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Advantages of ontology-driven approach• End users can focus more on exploring real-time datasets
• Semantics are explicitly specified and transferrable
• User interface allows domain ontologies to be interchangeable
24 |
Curation Coding Analysis, Monitoring, Management
Curation CodingAnalysis,
Monitoring, Management
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Fusing real-time events with domain knowledge
25 |
KnowledgeBase
Sensor NetworkReal-time data
Event of Interest
Query knowledge base(domain knowledge)
Notifications
e.g. Populate knowledge base with parameterised historical pipe failure data.
Infer likelihood of pipe failure based on physical attributes and known operating environment
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Modelling the feature of interest – pipe materials
26 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Event detections using dynamic and static info
28 |
> 200 PSI
+
Pipe material is PVCand
Risk level of pipe is A (good)
(Dynamic)
(Static)
Notification:
Location: Pipe XRisk of burst: LOW
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Event detections using dynamic and static info
29 |
> 200 PSI
+
Pipe material is PVCand
Risk level of pipe is E (bad)
(Dynamic)
(Static)
Notification:
Location: Pipe XRisk of burst: HIGH
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Summary
• Availability of real-time sensor data presents many potential applications
• Ontologies offer a means to capture semantics of a domain of discourse
• Ontology-driven approaches can assist user-definition of events over a given sensor network and consistent use of domain, application, sensor network semantics
• Shown how real-time events can be combined with domain knowledge for context sensitive event detection using ontologies
30 |
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Future work
Notification handling• Messaging queue systems• Attaching metadata based on
event rule semantics
More complex events• Event semantics • Incorporate processing-filters
User studies to evaluate the user interface
Deployments on actual sensor networks
31 |
A
BEvent
Smoothing function
Email / SMS
Database
Execute workflow
Existing alert systems
Ontology-enabledUser Interface
Sensor Network
Performing event detection over real-time sensor data using ontology-driven approaches | Jonathan Yu
Questions?
32 |
Land and WaterScott GouldResearch Projects Officert +61 3 9252 6103e [email protected] www.csiro.au/clw
ICT CentreKerry TaylorPrincipal Research Scientistt +61 2 6216 7038e [email protected] www.csiro.au/ict
Land and WaterDonavan MarneyResearch team leadert +61 3 9252 6585e [email protected] www.csiro.au/clw
LAND AND WATER
Thank youLand and WaterJonathan YuResearch Software Engineert +61 3 9252 6440e [email protected] www.csiro.au/clw
Land and WaterPaul DavisResearch Scientistt +61 3 9252 6310e [email protected] www.csiro.au/clw
Land and WaterBrad ShermanResearch Scientist
e [email protected] www.csiro.au/clw