train management platform for advanced maintenance of
TRANSCRIPT
Dept. of Information Technology – Broadband Communication Networks
Train Management Platform for advancedmaintenance of Passenger Information
Systems (PIS)
Stijn Verstichel ([email protected])A. Baart, G. Lievens, S. Latre, F. Deturck and F.Vermeulen
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 2
Introduction
Context : Trains
DynamicComplex
Problem : Maintenance
Preventive Replace at regular intervalsCorrective Replace when broken
Can be optimized
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 3
Outline
Introduction
From Idea towards Architecture
Ontology
Results and Conclusions
G10
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 4
Optimize PIS Maintenance
“We want to design an information-infrastructure, capable of correlating the available status information, in order to perform cost-effective
maintenance”
Current Work
Previous Research
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 5
Much Information Available
How ?
1. Perform measurements per componentOn demandScheduled
2. Store measurements3. Correlate the information4. Draw conclusions
Advantages !
RobustEfficient maintenanceLess expensive maintenancePrediction of faults and failuresSelf-descriptive….
}Use of Semantic Web’s First-Order Logic Ontology
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 6
From Idea to Architecture
Functional Building blocks
Obtain status-information from components
Analyze communication between components
Impose consequences
Provide feedback to the Users
Easily install now model of the system
Colelct information from imposed consequences
FunctionalBuilding blocks
s6G5
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 7
From Idea to Architecture
ScenariosObtain status-information from components
FunctionalBuilding blocks ScenariosFunctionalBuilding blocks Scenarios
Request Measurement
Request MeasurementRequest
Measurement
Result
ResultFilter
Store
s7G7
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 8
Communication
From Idea to Architecture
Architecture“topdown”-design method
FunctionalBuilding blocks Scenarios ArchitectureFunctionalBuilding blocks Scenarios ArchitectureFunctionalBuilding blocks Scenarios
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 9
From Idea to Architecture
Business Logic
MeasurementController
CommunicationRouter
ConsequencesController
CommunicationController
GUIAdapter
Data StorageRouter
Data StorageInterface
CommunicationInterface
WagonController
ArchitectureFunctionalBuilding blocks Scenarios
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 10
Data StorageController
Racer Local Mapper
Business LogicInterface
Reasoner B Local Inteface
RacerLocalInterface
RacerRemoteInterface
RacerRemote Mapper
Reasoner BLocal Mapper
ReasonerController
OntologyController
OntologyInterface Remote
Reasoner BRemote Mapper
Reasoner BRemote Interface
OntologyInterface Local
From Idea to Architecture
Data Storage
ArchitectureFunctionalBuilding blocks Scenarios
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 11
Introduction to ontology
RDF (Resource Description Framework)Basic expressiveness
OWL (Ontology Web Language)built on top of RDFLite: classification hierarchy and constraint featuresDL (Description Logic): maximum expressiveness without losing computational completeness and decidabilityFull: maximum expressiveness and syntactic freedom with no computational guarantees
Gre
ater
exp
ress
iven
ess
RDF
OWL
Lite
DL
Full
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 12
Key Features of Ontology
Self-descriptive
Application Logic inside the model
1st – order – logic
Automatic ValidationClassificationInference
Reasoner
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 13
Overview TIO.owl
PIS Ontology (Local)
hasMeasurement hasMeasurementLevel
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 14
PIS Ontology (Remote)
Overview TIO_remote.owl
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 15
Results
Maintenance Tool
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 16
Results
Maintenance Tool
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 17
Conclusions
Currently:
Maintenance = straightforward BUT inefficientMaintenance = f(MTBF)
failure
Proposed maintenance application
Maintenance optimized per componentMaintenance tool for visualisation
Maintenance = simple AND more efficientMaintenance = f(MTBF, cost)
Train Management Platform for advanced maintenance of Passenger Information Systems –Stijn Verstichel – Dept. of Information Technology – IBCN
p. 18
Conclusions
Value for MoneyOperator/Owner
+ Financially interesting+ Improved robustness
Maintainer+ Easier maintenance+ Improved planning
Passengers+ Greater uptime for components and systems+ Improved and up-to-date information
Ontology+ dynamic, reasoning, self-descriptive,…- complex and (currently) not yet performant