Delivering Digital Services Presented By: Eduard Gringinger Date:August 27, 2014.

Delivering DigitalServices

Presented By: Eduard Gringinger

Date: August 27, 2014

• Ontology-based representation & querying of DNOTAM– SWIM enabled component which you can add to any application

• Industrial research & experimental development project founded by the Austrian Research Promotion Agency (FFG)

• Started 2014, duration 3 years• Joint Undertaking of FREQUENTIS & University of Linz• Supported by AustroControl, Jet Aviation, FAA - FNS NDS &

Eurocontrol

Semantic DNOTAM

After closer evaluation of 389 NOTAMs, only 68 were applicable.

*presented by Ernie Bilotto during the NOTAM industry day

After fine-grade filtering only 32 NOTAMs affect the route of flight.

Meet the need from GA to Dispatcher, Pilots

8

More specific and relevant to route of flight

Filtered so pilots can prioritize

9

• Shorter PIB• Prioritize DNOTAM data by relative risk• PIB organization to aid readability

User expectations

10

• Capture implicit knowledge from OPS• Filtering rules are not known in full a priori• Generic AIXM event schema, independent of

DNOTAM event scenarios• Large set of filtering/prioritization rules

Challenges

11

Roadmap2014 20162015 2017

1

2

3

4

5

6

D3.1: Showcase Description

Requirement Analysis & Architecture

SemNOTAM Ontology, Rule and Query Language

SemNOTAM Knowledge Base

Prototype: SemNOTAM Knowledge Base

31-36

D2.1: Generic Concepts D2.4: Generic RulesQuery Language

D1.1: Requirements Specification

1-6

D1.2: System Architecture

7-12

D2.7: PrioritisationD2.8: Temporal Ordering

D2.9: Spatial Ordering

13-18 19-24 25-30

D2.2: TimeD2.3: Space

D3.5. Business Rules: D3.6: Query Templates:D3.7: ShowcaseRules Queries

D4.4.: Final

D5.4: SemNOTAM

D6.3 Final reportD6.6 Scientific Publications

D6.7 Final Website

Showcase Application

SemNOTAM ProtégéD4.2: Protégé Extension(D2.4, D2.5, D2.6)

D4.1: Protégé Extension

D6.1c: Minor

D5.3: SemNOTAM

D4.3: Protégé Extension(D2.7, D2.8, D2.9)

Web Query Interface

D6.5: Launch of website

Ontology Language Rule Language

D3.3:AIXM2Ontology MappingD3.4: Derived Concepts

Ontology

implementing D2.1, D2.2, D2.3

Mapper v1

Milestone Report Milestone Report

D5.1: AIXM-to-Ontology

D6.1b: Minor D6.2a: Major

D3.2: Primitive Concepts

D2.5: TimeD2.6: Space

Prototype: SemNOTAM Environment

Project Management & Dissemination

Milestone Report Milestone Report

Relevance, Importance Prioritisation, Spatialand Temporal Ordering

Ontology Framework Rules Framework Query Framework

D6.1a: Minor Milestone rep.D6.4: Project mgmt tools

D5.2: AIXM-to-Ontology

D6.2b: Major

Mapper v2

we are here

SWIMTechnical Infrastructure

Service Model

DataModel

Specific Service

SpecificData

SWIM, the Fantastic 3

AIRM

ISRM

TI

Digital Briefing

14

SWIMTechnical Infrastructure

Service Model

DataModel

Digital Briefing Service

WeatherService

FlightService

FlightData

AeronauticalFeatureService

AeronauticalData

WeatherData

Briefing Application

semNOTAM prototypes

15

Digital NOTAMAIXM5.1

GIS

SWIM

IntegratedDigital Briefing

METMaps

APT MapsED99

Aero. FeatureAIXM5.1 / SDD

semNOTAMPrioritization

Dispatcher Briefing

ePIB

Electronic Flight BagBriefing

16

class DigitalBriefing_SemNOTAM_Integration

AMISWMSAMISWMSAMISWMS

AIFSWFSAIFSWFSAIFSWFS

AIMSWMSAIMSWMSAIMSWMS

ePIBePIBePIB

SemanticNOTAMPrioritization



«participant»Integrated Digital

Briefing

AMISWMSAMISWMS

AIFSWFSAIFSWFS

AIMSWMSAIMSWMS

ePIBePIB




«participant»Digital NOTAM Prov ider

AIFSWFSAIFSWFS

«serviceInterface»AIFS WFS

+ getCapabilities()+ getFeaturesByFilter()

«messageType»Digital NOTAM Ev ent

tagsencoding = AIXM 5.1

«messageType»AIFS WFS Response

tagsencoding = HTTP

«messageType»AIFS WFS Request

tagsencoding = HTTP

«messageType»OGC Filter

tagsencoding = OGC Filter / CQL


«participant»Aeronautical Feature

Prov ider

AIFSWFSAIFSWFS

«messageType»AIXM Feature

tagsencoding = AIXM 5.1





«Participant»Semantic NOTAM

Prioritization Prov ider



AIFSWFSAIFSWFS

«serviceInte...Semantic NOTAM

Prioritization

«messageType»Serv ice NOTAM

Prioritization Request

«messageType»Serv ice NOTAM

Prioritization Response

«messageT...Flight Plan

tagsencoding = FIXM

Trajectory

Flight Phase

Trajectory Point

Trajectory Segment

Note

Digital NOTAM Ev ent Reference

AIFS

AMIS

AIFS

AIMS

ePIB /HMI

AIFSWFS

Semantic NOTAMPrioritization

Semantic NOTAMPrioritization

AIFS

1

0..*

0..*

«trace»

0..*

1..*

«expose»

«expose»

«expose»

1..*

1

1

soaml DigitalBriefing_SWIM_TI_SemNOTAM_Integration

SWIM Node 14.02.09


«participant»Digital NOTAM

Prov ider

AIFSWFSAIFSWFS


«participant»Aeronautical Feature

Prov ider

AIFSWFSAIFSWFS

AMISWMSAMISWMSAMISWMS


AIMSWMSAIMSWMSAIMSWMS




«participant»Integrated Digital

Briefing

AMISWMSAMISWMS

AIFSWFSAIFSWFS

AIMSWMSAIMSWMS







«Participant»Semantic NOTAM

Prioritization Prov ider



AIFSWFSAIFSWFS

AIFS

AIFSWFS

AMIS

AIFS

AIMS

Semantic NOTAMPrioritization Semantic NOTAM

Prioritization

AIFS

«messageType» Service NOTAM Prioritization Request,«messageType» Service NOTAM Prioritization Response

«messageType» AIMS WMS Request,«messageType» AIMS WMS Response

«delegate»«serviceChannel»

«messageType» AIFS WFS Request,«messageType» AIFS WFS Response


«messageType» AIFS WFS Request,«messageType» AIFS WFS Response


«messageType» AMIS WMS Request,«messageType» AMIS WMS Response

• Protégé– OWL concepts: Classification of DNOTAMs– SWRL: Filtering rules– SWQRL: Closed-world queries

• Business Rule Engines– Oracle Business Rules– Drools

• KR-languages and ontologies– Datalog (DLV)– F-logic (Flora-2, ObjectLogic)

Technology Choices

17

• Approach– OWL concepts: Classification of DNOTAM– SWRL: Filtering rules– SWQRL: Closed-world queries

• Implemented mini-prototype– Representation of selected AIXM elements in OWL– Demonstrates conceptual feasibility of approach

OWL/SWRL/SQWRL

18

• Approach– DNOTAMs correspond to business cases– Filtering rules correspond to business rules

• But notice the following mismatch:– BREs usually support business cases with simple

structure and multiple types of BRs– DNOTAMs have complex structure of a single type

Business Rule Engines (BREs)

19

• Datalog: DLV– No support of complexity management in KR: Flat

predicates (leads to n-ary predicates with large n), no abstraction capabilities (later additions require change of arity of predicates)

– Limited extension facilities• F-logic

– Frames (complex object structures)– ObjectLogic and Flora-2 come with extension support

KR-Languages

20

• Conceptually Protégé and KR-languages suitable

• BREs with conceptual mismatch

Summary

21

Knowledge Base Overview

OWL Open Standard Languages

SemNOTAM Language Extensions (Work Package 2)

SemNOTAM Domain Knowledge

Ontology Rule Query

SemNOTAM „Data“

SQWRL SWRL

+SemNOTAM generic concepts +AIXM time +AIXM space

+SemNOTAM generic rules +AIXM temp. rel. +AIXM spatial rel.

+prioritisation +temporal ordering +spatial ordering +prioritisation

Derived Concepts (defined upon Primitives)

Primitive Concepts (1:1-AIXM-Correspond.)

Business Rules - Relevance - Importance

Query Templates - prioritisation - spatial, temporal ordering

Baseline Data

Situation Specific Data

NOTAMs

(time, airplane characteristics, airplane position, flight phase)

use use

uses uses

Legend:

exists

transformed existing knowledge

new knowledge (otherwise in program code)

(Work Package 3)

22

• Data Preparation– Prepare data such that subsequent filtering effort can be

reduced• Pre-filtering

– Coarse filtering in order to reduce DNOTAMs set• Filtering

– Relevance: discard irrelevant DNOTAMs– Priority: e.g. scenario-based, for a specific flight phase

Filtering & Prioritization

23

DNOTAM Rule Set

24Number of Rules

Rule Set Complexity

Relevant

Noise

Result Set

Rule Classification

25

Secondary Rules

Secondary Rules

Key Rules

Rule Complexity

Rule Selectivity

low

high

low high

• Acquire more operational input• Involve pilot associations

• Integration of FIXM 3.0• Mini Global Demo 2015

• Dispatcher briefing prototype• EFB prototype afterwards

Future Steps

Questions

27

Contact Information

[email protected]

semNOTAM.frequentis.com

28

mailto:[email protected]

Delivering Digital Services Presented By: Eduard Gringinger Date:August 27, 2014.

Documents

business cases filtering

airm isrm ti slide

filtering rules swqrl

ops filtering rules

dnotam data

prefiltering coarse

grade filtering

approach dnotams