Download - SSG4Env EGU2010
Speaker:
SemsorGrid4Env: Semantic Sensor Grids for
Rapid Application Development forEnvironmental Management
FP7-223913
European Geosciences Union 2010
From Sensors to Interoperable Sensor Networks
Vienna, 6th May 2010
Jean-Paul Calbimonte,Universidad Politécnica de Madrid semsorgrid4env.eu
Table of Contents
• The Consortium• Project Challenges and Main Outcomes• Project Plan & milestones• Highlights
• Architecture• Data management• Registries• Semantic Integration• Application Tier
EGU 2010 - Vienna, 6 May 2010 2
The Team
1. Universidad Politécnica de Madrid, (UPM,
Spain)
2. University of Manchester (UNIMAN, UK)
3. National and Kapodistrian University of
Athens (NKUA, Greece)
4. University of Southampton (SOTON, UK)
5. Deimos Space SLU (DMS, Spain)
6. EMU Ltd. (EMU, UK)
7. TechIdeas (TI, Spain)
3
3
3
1
3EGU 2010 - Vienna, 6 May 2010
Project Challenges
Integrated information space •Discovery new sensor networks•Integrate with existing ones
• Integrate possibly other data sources (e.g., historical databases)
020406080
100
1ertrim.
3ertrim.
Este
Oeste
Norte
sens or networks
legacy data sources
semantic data integration and querying
thin applications (mashups )
regis tries
middleware
Rapid development • flexible and user-centric decision
support systems • Use data from multiple
autonomous independently deployed sensor networks and other applications.
4EGU 2010 - Vienna, 6 May 2010
Main Outcomes (I)
System Level (WP1)• An architecture for the design and implementation of open large-
scale Semantic Sensor Grids.• A reference SemsorGrid4Env implementation instantiating the
architecture
Component-level (WP2-WP5):• New techniques and tools for semantic-based data management
over the heterogeneous data streams that stem from autonomously deployed sensor networks. (WP2)
• Scalable and fault-tolerant resource discovery mechanisms for sensor registries. (WP3)
• The semantic infrastructure (including ontologies) needed to facilitate the integration of data coming from heterogeneous and distributed sensor networks, legacy databases and applications. (WP4)
• Higher-level application programming interfaces that ease the rapid generation of thin applications (e.g., mashups) of data from sensor networks and historical databases. (WP5)
Two environmental management applications (WP6-WP7)
5EGU 2010 - Vienna, 6 May 2010
Main Outcomes (II)
Fire Risk Monitoring and Warning in Spain
(technology-driven)
Coastal and Estuarine Flood Warning in Southern UK.
(established early adopter community)
MIDDLEWARE
THIN APPLICATIONS
Sensor network 1 Sensor network 2
EARTH OBSERVATION PRODUCTS
-1
-0.5
0
0.5
1
1.5
2
2.5
3
40
03
40
04
40
05
40
06
40
07
40
08
40
09
40
10
40
11
40
12
T im e sin c e sta r t o f ye a r (h r )
Sta
ge
(m
)
M o de l p re d ic tion s M e a su rem en t s R es idu a ls
W W W d a ta a cce ss
N a tio n a l S u rge M o d e l
R e g ion a l tid e /wa v e /m e t d a ta
F ig . X L iv e - tim e d a ta a ss im ila tio n flo o d m o d e llin g : M u ltip le d a ta a cce ss
S o le n t R e -de p lo ya b le G S M -b a sed se n so r n e two rk
G S M
6EGU 2010 - Vienna, 6 May 2010
7
Why SSG4Env?
• Flood and fire have significant environmental and economic impact in Europe
• Significant potential from emerging technologies to assist users by:• Improved monitoring by deployed & emerging sensor networks• New capabilties in data integration including live data streams• Rapid development of flexible and user-centric decision support
systems • Semantic Web supporting discovery and integration
• SSG4Env combines expertise and technology in all of these areas to provide solutions which are simple, live and dynamic
EGU 2010 - Vienna, 6 May 2010
8
Workpackage Structure and Deliverables
D1.1: Setup of software development technologies
D1.2: Deployment of technological infrastructure
D1.3: SemsorGrid4Env Architecture
D2.1: Data Requirements, Data Management and Analysis Issues and Query-Based Functionalities
D3.1: Data models and languages for registries in SemsorGrid4Env
D3.2: Distributed data structures and algorithms for a Semantic Sensor Grid registry
D4.1: Design of the SemsorGrid4Env ontology-based data integration model
D5.1: Specification of high-level application programming interfaces
D6.1: Requirements specification
D6.2: Deployment of the sensor network
D7.1: Requirements specification
D7.2: Deployment of the FloodNet sensor network in the Solent
D8.1: Quality and Risk Contingency Plan
D8.2: Gender Action Plan
D9.1: SemsorGrid4Env Website
D9.2: Plan for Dissemination Activities
D9.3:SWOT Analysis
8EGU 2010 - Vienna, 6 May 2010
Main Project Phases
Phase 1Months 1-6(Specification)
1. Induction, know-how, and use case gathering. 2. WP 1-5 will collaborate to assimilate the current architectural and technological
features and challenges of all the areas involved; and identify appropriate approaches for their integration.
3. WP 1-5 work with WP 6-7 to develop use case specifications that will form the drivers and the evaluation framework for the rest of the project
Phase 2Months 7-22(Development 1)
1. A specification of the architecture incorporating all services from WP 2-5.2. A first specification of the use case applications3. A first prototype of the individual systems and of the integrated middleware
Phase 3Months 23-34(Development 2)
1. A refined specification of the architecture incorporating all services from WP 2-5.
2. A second specification of the use case applications3. A second prototype of the individual systems and of the integrated middleware
Phase 4Months 33-36(Final Evaluation)
1. Evaluation of individual systems and integrated middleware.2. Evaluation of use cases.
9EGU 2010 - Vienna, 6 May 2010
WP1: SSG4Env General Architecture
10
Properties:• Any service may directly call
any other service.• Pre-existing services may be
called by any service.• Independent development of
services.• Based on WS-* standards: WS-
RF, WS-DAI, WS-N.
Key features:• Data tier services wrap
concrete data resources.• Semantic middleware adds
value to services in application and data tiers.
EGU 2010 - Vienna, 6 May 2010
• Principal outcomes:• SNEE query processor / SNEEql query language• Documented requirements from use cases, to the level
of queries and data analyses.• Support QoS-aware evaluation within in network query
optimizer.• Developed out-of-network query compiler and evaluator
from to support integration queries.
WP2: Data & Stream Query Processor
SELECT RSTREAM t.id, w.speed, w.dirnFROM wind[NOW] w, tree[NOW] tWHERE t.smoke > 0AND sqrt((t.locx - w.locx)^2 + (t.locy - w.locy)^2) <= 40
EGU 2010 - Vienna, 6 May 2010 11
• Defined the data model stRDF and the query language stSPARQL, based on the paradigm of constraint databases.
• Represent thematic and spatial metadata that change over time. Coupled with the RDFS/OWL ontologies of WP4.
• Developed a formal semantics and algebra for stSPARQL on which we base our implementation.
• Development of Strabon: a centralized implementation of a subset of stSPARQL.
WP3: Semantic Registry
EGU 2010 - Vienna, 6 May 2010 12
• Designed, implemented and deployed a Semantic Integration Service • Extend existing ontology-based data integration models
to take into account sensor networks streaming data, semantic heterogeneity and quality of service
• Specified a suite of sensor network ontologies that will be used for describing sensors and related data for the SemSorGrid4Env software architecture
WP4: Semantic Infrastructure
EGU 2010 - Vienna, 6 May 2010 13
Query reconciliation
q qrQuery
canonisation
Qc
Distributed Query
Processing
Data decanonisation
Data reconciliationd dr
Dc
Clie
nt
O-O mapping S2O mappings
SPARQLSTR (Og) SPARQLSTR (O1 O2 On) SNEEql (S1 S2 Sn)
SNEEql’ (S1 S2 Sn)
[tuplel1 l2 l3][tripleO1 O2 On][tripleOg]
Semantic Integrator
WP2: Semantic IntegratorOntology-based data access
EGU 2010 - Vienna, 6 May 2010 14
SSG4Env Application Tier (WP 5, 6 & 7)
• High level API provides functionality for domain developers (API) and domain users (web apps)
• Supports applications in WP6 and WP7
• Resource-centric including Linked Data
• Embraces and investigates interplay of SOA and ROA
15EGU 2010 - Vienna, 6 May 2010
Summary (I): Application highlights
• Application requirement specifications• Sensor deployment in the UK Solent area• Early mashup developments for flood warning
• In order to engage more quickly potential users and other stakeholders.
16EGU 2010 - Vienna, 6 May 2010
Summary (II): Technical highlights
• Integration platform (WP1)• Architecture, validated with the application use cases (WP1)• Selection of outlier detection algorithms (WP2)• Out-of-network event stream query processor (WP2)• TinyOS2 code generator for the in-network SNEE (WP2)• Spatio-temporal extension of SPARQL (stSPARQL) (WP3)• Ontology-based streaming data access (WP4)• Selection of ontologies to be reused (WP4)• API combining RESTful and Linked Open Data approaches
(WP5)• A proposal for the identification, naming and generation of
Linked Stream Data (WP5)
17EGU 2010 - Vienna, 6 May 2010
Speaker:
SemsorGrid4Env: Semantic Sensor Grids for
Rapid Application Development forEnvironmental Management
FP7-223913
European Geosciences Union 2010
From Sensors to Interoperable Sensor Networks
Vienna, 6th May 2010
Jean-Paul Calbimonte
EGU 2010 - Vienna, 6 May 2010
• Mashups provide rapid development of web interfaces to support custom requirements.
• Mashups require combined access to:• Sensed data from multiple sensors.• Stored data from multiple sources.• Ontologies for linking independent sources.
• The aim of the architecture is to deliver appropriate abstraction and integration services for the mashups.
Top Level Requirements
EGU 2010 - Vienna, 6 May 2010
The Consortium Classified
Seven partners• Four universities• 2 SME• 1 large company
Four major sectors• Education• IT• Aerospace Engineering• Environment
Technological core competencies• Sensor Networks (UNIMAN,
SOTON-ECS, NKUA)• Semantics (UPM, UNIMAN,
SOTON-ECS)• Grid (UNIMAN, TI, SOTON-
ECS, UPM)• P2P (NKUA)• Rapid Application
Development (SOTON-ECS)
Use Cases• Flood warning (EMU,
SOTON-GEODATA)• Fire warning (DMS)
21EGU 2010 - Vienna, 6 May 2010
conceptmap-def WindSpeedMeasurement
uri-as
concat('ssg4env:WindSpeedMeasurement_',
windsamples.sensorid,windsamples.ts)
described-by
attributemap-def hasSpeed
operation "constant"
has-column windsamples.speed
dbrelationmap-def isProducedBy toConcept Sensor
joins-via
condition "equals"
has-column sensors.sensorid
has-column windsamples.sensorid
conceptmap-def Sensor
uri-as
concat('ssg4env:Sensor_',sensors.sensorid)
described-by
attributemap-def hasName
operation "constant"
has-column sensors.sensorname
Measurement
WindSpeedMeasurement
Sensor
isProducedBy
hasName xsd:string
hasSpeed xsd:float
S:WindSamples
- ts- speed- direction- sensorid
T:Sensors
- sensorid- sensorname
S2O: Mapping streams to ontologies
EGU 2010 - Vienna, 6 May 2010
PREFIX fire: http://www.semsorgrid4env.eu#
PREFIX rdf: http://www.w3.org/1999/02/22-rdf-syntax-ns#
SELECT ?speed ?name
FROM STREAM <http://www.ssg4env.eu/Readings.srdf>
[RANGE 10 MINUTE STEP 1 MINUTE]
WHERE {
?WindSpeed a fire:WindSpeedMeasurement;
fire:hasSpeed ?speed;
fire:isProducedBy ?sensor;
fire:hasTimestamp ?time.
?sensor a fire:Sensor;
fire:hasName ?name.
}
SELECT concat(‘ssg4env.eu#WindSpeedMeasurement' , windsensor.id, windsensor.ts ) as a1 , concat( ‘ssg4env.eu#Sensor' , sensors.sensorid ) as a2 FROM sensors, windsensor[ FROM NOW - 10 TO NOW MIN] WHERE ( sensors.sensorid = windsensor.id )
Semantic Integrator
Work in progress: removing redundant queries, basic optimisations, more complex scenarios
Transforming SPARQLSTR to SNEEql