Knowledge-Based Engineering · Technical Specification Statement of Work Logistics Support ISO...
Transcript of Knowledge-Based Engineering · Technical Specification Statement of Work Logistics Support ISO...
Knowledge-Based
Engineering
Elena Gallego. Consulting Director at TRC.
All rights reserved © The REUSE Company 2019
Pentagon wars
All rights reserved © The REUSE Company 2019
“◈ It is the idea of knowledge digitalization
one of the challenges that organizations
need to pursue in order to build better
systems or deliver better services, in less
time, with less money and more
efficiency.
3
All rights reserved © The REUSE Company 2019
Knowledge Base Purpose
◈ A few goals of the Knowledge Based Systems are:
⬥ To ensure that the Product or Services development main documents
are Complete, Consistent and Correct
⬥ To support the quality analysis of requirements, models, and
unstructured information or data
⬥ To reuse the Organizational or Domain information among several
projects
⬥ To share knowledge between the different stakeholders in the process
⬥ To infer behavioral patterns from legacy assets
4
All rights reserved © The REUSE Company 2019
Knowledge Framework
5
Knowledge
Management
Traceability
AuthoringQuality
Management
Users
Knowledge Base
All rights reserved © The REUSE Company 2019
40% Cost
Saving(average value)
Constructive Systems Engineering Model - COSYSMO
6
All rights reserved © The REUSE Company 2019
The REUSE Company, deploying services and systems engineering from experience
Innovative technologies applied
to
Knowledge Reuse
7
All rights reserved © The REUSE Company 2019
Product
Oriented
Knowledge Based System: Infers and Uses a Knowledge Base to solve complex problems
Services
Oriented
8
All rights reserved © The REUSE Company 2019
Product
Oriented
Knowledge Based System: Infers and Uses a Knowledge Base to solve complex problems
Services
Oriented
9
01Controlled Organizational and
Project Vocabulary for a common
understanding among stakeholders
Vocabulary
02Recreate and capture the
system architectures
represented in views and
models. Stablish
relationships among system
and system elements
Architectures
04Representation of
assets semantic through
SRL – System
Representation
Language
Formalization03
Represent requirements
similarities and enable formal
representation, automatic
recognition and aid authors
Patterns
05A combination of rules,
tasks and groups to
infer information from
valuable assets
Reasoning
All rights reserved © The REUSE Company 2019
Product Oriented
◈ Product Breakdown
Structures
◈ Architectures
◈ Thesaurus
◈ Patterns
◈ Controlled Vocabulary
Knowledge Organization
Services Oriented
◈ Unstructured data
◈ Unknown inputs of
information
◈ Thesaurus
◈ Patterns
◈ Controlled Vocabulary
10
All rights reserved © The REUSE Company 2019
Different Knowledge Base Applications
Procurement
Product Line
Predictive Maintenance
11
Services
Products
All rights reserved © The REUSE Company 2019
Procurement
Knowledge
Base
12
All rights reserved © The REUSE Company 2019
http://ec.europa.eu/internal_market/scoreboard/performance_per_policy_area/publi
c_procurement/index_en.htm 13
All rights reserved © The REUSE Company 2019
Procurement Knowledge Base Objectives
◈ Automatic Analysis of the Evaluation Criteria
⬥ Traceability between user and business needs and evaluation results
◈ Ensure complete documentation development process
⬥ Guarantee stakeholder priorities are covered at different documentation
levels
◈ Analysis of life-cycle cost among involved areas
⬥ Completeness and Consistency at all levels
◈ Support requirements and models development
⬥ Patterns guidance at all levels
14
All rights reserved © The REUSE Company 2019
Technical Specification Statement of Work Logistics Support
ISO 29148CompletenessConsistency
Correctness Completeness Consistency
ISO 15288 | MIL-Hdbk-245
Correctness Completeness Consistency
ASD SX 000iJSP886 | Def Stan 00-600
Correctness Completeness Consistency
ISO 29148
Procurement Knowledge Base
15
All rights reserved © The REUSE Company 2019
Quality in context
Completeness
Consistency
Correctness
Completeness
Consistency
16
All rights reserved © The REUSE Company 2019
Correctness Quality Check
◈ The Correctness Quality
Set:
⬥ Characteristics coverage
⬥ Ontology dependency
⬥ Effort needed to fix identified
error
17
All rights reserved © The REUSE Company 2019
Completeness Quality Check
Specification viewpoint Ontology viewpoint
18
All rights reserved © The REUSE Company 2019
Consistency Quality Check
Traces Inconsistency
Specification Inconsistency Ontology Inconsistency
19
All rights reserved © The REUSE Company 2019
PROCUREMENT
KNOWLEDGE BASE
VOCABULARY PATTERNS & FORMALIZATION QUALITY METRICS
• TERMINOLOGY
• NLP RULES
• SCM VIEWS
• CLUSTERS
• PATTERNS
• PATTERN GROUPS
• PROPERTIES
• RELATIONSHIPS
• CORRECTNESS RULES
• COMPLETENESS RULES
• CONSISTENCY RULES
• QUALITY & AUTHORING CONFIGURATIONS
Is formed by
Is formed by
Is formed by
Procurement Knowledge Base main Knowledge work-products
21
All rights reserved © The REUSE Company 2019
Procurement Knowledge Base Organization
22
SSS
Library
CLS/ILS
Library
SoW
Library
PROCUREMENT
KNOWLEDGE BASE
VOCABULARYPATTERNS &
FORMALIZATIONQUALITY METRICS
• TERMINOLOGY
• NLP RULES
• SCM VIEWS
• CLUSTERS
• PATTERNS
• PATTERN
GROUPS
• PROPERTIES
• RELATIONSHIP
S
• CORRECTNESS
RULES
• COMPLETENESS
RULES
• CONSISTENCY
RULES
• QUALITY &
AUTHORING
CONFIGURATIONS
Is formed by
Is formed by
Is formed by
Procurement Knowledge Base
All rights reserved © The REUSE Company 2019
Procurement Knowledge Base Organization
◈ The Procurement Knowledge Base has sub-Knowledge Bases
to cover the three main sort of documents for each of the
knowledge work products.
23
SSS
CLS/ILS
SoW
Each database is developed
by different teams, covering
different content
All rights reserved © The REUSE Company 2019
Knowledge Base Libraries
24
INCOSE
• Guide for Writing
Requirements
EARS
• Requirement Patterns
ECSS
• Domain Knowledge
ISO 26262
• Domain Knowledge
Procurement
• Know-how for Automatic Evaluation
All rights reserved © The REUSE Company 2019
Product Line
Engineering
Knowledge
Base
25
All rights reserved © The REUSE Company 2019
The European automotive sector differed greatly in the level of variety they offered to
customers, although variety had little relation to unitary sales.
BodiesPower
trainsTotal number of variations
European
units sales in
2002
Mercedes E-Class 30 15 3,347,807,348,000,000,000,000,000 157,584
BMW 3-Series 10 20 64,081,043,660,000,000 350,723
Peugeot 206 5 24 1,739 596,531
26
Source: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.469.2061&rep=rep1&type=pdf
All rights reserved © The REUSE Company 2019
Product Line Knowledge Base Objectives
◈ Extraction of requirements from the product-line assets
⬥ Automatic allocation of assets from the solution space to requirements
document
◈ Identification of the variant features in the requirements
⬥ Patterns and Thesaurus to cover commonality and variability
◈ Coverage of the specific system features
⬥ Knowledge interfaces with Product Lifecycle Management software tools
27
All rights reserved © The REUSE Company 2019
The Product Line Knowledge Base is formed by all types of Knowledge
28
Stakeholder
needs and
requirement
s
OperationBusiness or
Mission
Analysis
Maintenanc
eDisposal
System
requirement
s
Architecture
Design
Validation
Transition
Verification
System
AnalysisIntegration
Implementation
Knowledge Base
All rights reserved © The REUSE Company 2019
“◈ The goal of this process is to automatically
generate the right set of patterns to formalize the
features of the products and the different
semantics and values from the natural language
into SRL and later use the formalization of the
different products to generate the variability
model based on the captured features.
29
All rights reserved © The REUSE Company 2019
Operating between assets
Knowledge
QualityModels
Requirements
All rights reserved © The REUSE Company 2019
Ground Operational Amp
<<Connection>
>
Capacitor
<<Connection>
>
Resistor
Operational Amp
Resistor
<<Connection>
>
<<Connection>
>
Operational Amp
Operational Amp
Common Representation Language
Battery<<Message Sending>>
Power Control
SystemCharge System
<<Message
Sending>>
Power Control
System
All rights reserved © The REUSE Company 2019
A simple example of a pattern matching, and relationships generation
When switched on, the Cab radio shall operate within a temperature range of -20°C
to +70°C
When [TRIGGER] and [PRECONDITION], the [ACTOR] shall [ACTION] [OBJECT]
32
Temperature
-20°C +70°C
Operation
Cab radio Temperature
All rights reserved © The REUSE Company 2019
Predictive
Maintenance
Knowledge
Base
33
All rights reserved © The REUSE Company 2019
The Cost of Operations & Maintenance
34
6016
12
12
O&M Tax Management Insurance
85
15
Wind Turbines Rest of the Wind Farm
55
25
20
Spare parts Personnel Consumables
Exploitation Costs O&M Costs Wind turbine Costs
All rights reserved © The REUSE Company 2019
“◈ Combination of a predictive and
preventive maintenance with an efficient
corrective action can save up to 70%
wind turbine operational costs.
35
All rights reserved © The REUSE Company 2019
Predictive Maintenance Knowledge Base Objectives
◈ Automatic failures diagnosis through condition monitoring
⬥ Reasoning engine to identify deviations in data
◈ Representation of legacy data
⬥ Knowledge interface with the data
◈ Predictive techniques to identify behavioural patterns
⬥ Definition of patterns that capture data variations
36
All rights reserved © The REUSE Company 2019
Data MiningKnowledge
Base
Infer data structures
37
Data
Information containers
Information
Set of patterns
All rights reserved © The REUSE Company 2019
93.77%Knowledge Based systems can infer the status of a wind turbine from its
measurements and predict its behaviour in 93.77% of the cases
38
All rights reserved © The REUSE Company 2019
Efficient
application of
Knowledge-
Based
Engineering
39
All rights reserved © The REUSE Company 2019
Make it SimpleProcess automation and User’s Support
40
Systems Engineering Handbook. INCOSE-TP-2003-002-2015 Systems Engineering Handbook. INCOSE-TP-2003-002-2015
All rights reserved © The REUSE Company 2019
Knowledge
Based Systems
to solve complex
problems
Knowledge
Digitalization to
improve the way
we do
engineering
And all this to conclude…
Define the right
strategy to
design and
implement the
Knowledge
Base
41
All rights reserved © The REUSE Company 2019
Thank you!Any questions?
You can reach me at [email protected]
42