Dynamically Confjgurable Traceability Semantics Tarski: A Platform - - PowerPoint PPT Presentation

dynamically confjgurable traceability semantics tarski a
SMART_READER_LITE
LIVE PREVIEW

Dynamically Confjgurable Traceability Semantics Tarski: A Platform - - PowerPoint PPT Presentation

Introduction Geylani Kardas 4 F. Erata et al. 3rd Workshop on Dependability at Izmir Institute of Technology 5 Academy Directorate, HAVELSAN Inc., Ankara, Turkey 4 International Computer Institute, Ege University, Izmir, Turkey 3 System


slide-1
SLIDE 1

Introduction Approach Demonstration Conclusion and Future Work

Tarski: A Platform for Automated Analysis of Dynamically Confjgurable Traceability Semantics

Ferhat Erata1,2 Moharram Challenger1,4 Bedir Tekinerdogan1 Anne Monceaux3 Eray Tuzun5 Geylani Kardas4

1Information Technology Group, Wageningen University, The Netherlands 2UNIT Information Technologies R&D Ltd., Izmir, Turkey 3System Engineering Platforms, AIRBUS Group Innovations, Toulouse, France 4International Computer Institute, Ege University, Izmir, Turkey 5Academy Directorate, HAVELSAN Inc., Ankara, Turkey

3rd Workshop on Dependability at Izmir Institute of Technology

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-2
SLIDE 2

Introduction Approach Demonstration Conclusion and Future Work

Acknowledgements

Scientifjc and Technological Research Council of Turkey (TUBITAK), Technology and Innovation Funding Programs Directorate (TEYDEB) under project# 9140014, 9150181 Minister for the Economy, Industry and Digital Afgairs of France, Directorate-General for Enterprise (DGE) under contract# 14293020 European Cooperation in Science and Technology (COST) Action IC1404 ”Multi-Paradigm Modelling for Cyber-Physical Systems”

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-3
SLIDE 3

Introduction Approach Demonstration Conclusion and Future Work

Exploitations

ITEA-ModelWriter: Synchronized Document Engineering Platform https://itea3.org/project/modelwriter.html ITEA-ASSUME: Afgordable Safe & Secure Mobility Evolution https://itea3.org/project/assume.html Source codes, datasets and screencasts are available at: https://github.com/ModelWriter/WP3

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-4
SLIDE 4

Introduction Approach Demonstration Conclusion and Future Work

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-5
SLIDE 5

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-6
SLIDE 6

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

What is Traceability? Traceability can be defjned as the degree to which a relationship can be established among work products (aka. artefacts) of the development process. What is case-based or project-based traceability confjguration? Rigorously specifjcation the semantics of traceability elements. Why is Reasoning about Traceability important? Richer and precise automated traceability analysis. Compliance and Certifjcation in automotive and aviation industries.

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-7
SLIDE 7

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

Challenges of Traceability in Industry

Semantically meaningful traceability traceability relations should have a rich semantic (meaning) instead of being simple bi-directional referential relation Confjguration of traceability (possibly dynamically) Traceability Semantics is often statically defjned. The semantics cannot be easily adapted for the needs of difgerent projects. Difgerent traceable elements and the relation types exist in industrial settings, Likewise, difgerent traceability analysis scenarios exists. Several industries demands formal proofs of Traceability.

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-8
SLIDE 8

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

Challenges of Traceability in Industry

Semantically meaningful traceability traceability relations should have a rich semantic (meaning) instead of being simple bi-directional referential relation Confjguration of traceability (possibly dynamically) Traceability Semantics is often statically defjned. The semantics cannot be easily adapted for the needs of difgerent projects. Difgerent traceable elements and the relation types exist in industrial settings, Likewise, difgerent traceability analysis scenarios exists. Several industries demands formal proofs of Traceability.

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-9
SLIDE 9

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

Challenges of Traceability in Industry

Semantically meaningful traceability traceability relations should have a rich semantic (meaning) instead of being simple bi-directional referential relation Confjguration of traceability (possibly dynamically) Traceability Semantics is often statically defjned. The semantics cannot be easily adapted for the needs of difgerent projects. Difgerent traceable elements and the relation types exist in industrial settings, Likewise, difgerent traceability analysis scenarios exists. Several industries demands formal proofs of Traceability.

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-10
SLIDE 10

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

Challenges of Traceability in Industry

Semantically meaningful traceability traceability relations should have a rich semantic (meaning) instead of being simple bi-directional referential relation Confjguration of traceability (possibly dynamically) Traceability Semantics is often statically defjned. The semantics cannot be easily adapted for the needs of difgerent projects. Difgerent traceable elements and the relation types exist in industrial settings, Likewise, difgerent traceability analysis scenarios exists. Several industries demands formal proofs of Traceability.

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-11
SLIDE 11

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-12
SLIDE 12

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

Airbus Group Innovations

System Installation Design Principles

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-13
SLIDE 13

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

Airbus Group Innovations

System Installation Design Principles

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-14
SLIDE 14

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

Airbus Group Innovations

System Installation Design Principles

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-15
SLIDE 15

Introduction Approach Demonstration Conclusion and Future Work Motivation Industrial Use Cases

Havelsan Aerospace Electronics Industry

Application Lifecycle Management

DO-178C Software Considerations in Airborne Systems and Equipment Certifjcation Traceability DO-178 requires a documented connection (called a trace) between the certifjcation artifacts. For example, a Low Level Requirement (LLR) traces up to a High Level Requirement (HLR). A traceability analysis is then used to ensure that each requirement is fulfjlled by the source code, that each requirement is tested, that each line of source code has a purpose (is connected to a requirement), and so

  • forth. Traceability ensures the system is complete.
  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-16
SLIDE 16

Traceability Analysis Activities defjned in DO-178

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-17
SLIDE 17

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-18
SLIDE 18

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

A conceptual model for traceability and its extension

1 .. * {ordered}

Trace-element Trace-link Trace-location

0..*

FileLocation TextLocation JavaLocation XMILocation depend require satisfy contain ... generate ContractRequirement SystemRequirement ModelElement ... Traceability Information

{unique} 1..* source

...

{xor} 0..1 targets elements

Requirement EClassifier

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-19
SLIDE 19

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Semantics of contain relation (represents decomposition)

1 .. * {ordered}

Trace-element Trace-link Trace-location

0..*

FileLocation TextLocation JavaLocation XMILocation depend require satisfy contain ... generate ContractRequirement SystemRequirement ModelElement ... Traceability Information

{unique} 1..* source

...

{xor} 0..1 targets elements

transitive, irreflexive, antisymmetric, injective relation

Requirement EClassifier

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-20
SLIDE 20

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Semantics of ContractRequirement

1 .. * {ordered}

Trace-element Trace-link Trace-location

0..*

FileLocation TextLocation JavaLocation XMILocation depend require satisfy contain ... generate ContractRequirement SystemRequirement ModelElement ... Traceability Information

{unique} 1..* source

...

{xor} 0..1 targets elements

Requirement EClassifier

Disjoint subset of the set of Requirement, Domain and Co-domain of contain relation

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-21
SLIDE 21

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-22
SLIDE 22

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Fragments of a traceability instance

e3: EClass e2: EClass f1: Field m1: Method c2: Class ECore Design Document [Y] Java Package [Z] s1: SystemReq s2: SystemReq Requirement Document [X] {(E1,E3)} {(E2,E1)} {(E3)} {(U1)} {(E2)} {(S1)} {(S2)} {(E1)} { (F1, {(S1, E1)}) } = { (F1, S1), (F1, E1) } = jre1: {(F1, S1, E1)} { (U1, { (S1), (S2) }) } m2: Method c3: Class {(M2)} {(C3)} Java Package [W] ej1: {(E2, M1)} ej3: {(E3, P2)} {(C3,M2)} W = {(P2), (C3), (M2), (C3,M2)} Y = {(E1), (E2), (E3), (E2,E1), (E1,E3)} X = {(U1), (S1), (S2), (U1,S1), (U1,S2)} {(C1,M1)} {(C1,F1)} {(C1)} {(C2)} {(F1)} {(M1)} Z = {(P1), (C1), (C2), (F1), (M1), (C1,F1), (C1,M1)} + W Trace-Relations = {(S1), (E1), (E2), (E3), (F1), (M1), (E2,M1), (E2,M2), (E3,P2), (F1,S1,E1), (P2)} Formalization = ec: E ->(C + M + F) and ep: E -> P and fse: F -> S -> E ej2: {(E2, M2)}

<<refines>>

u1: UserReq

<<EReference>>

e1: EClass

< < i n h e r i t s > > <<calls>>

c1: Class {(P2)} {(P1)}

f1.source jre1.targets[1] jre1.targets[0]

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-23
SLIDE 23

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

First-order relational model of the traceability instance

The universe of traceability of the current state DT : {S1, E1, E2, E3, F1, M1, M2, P2} The type signature ΣT : {REJ ⊑ E → C ⊔ M ⊔ F, RJRE ⊑ F → S → E} The relational model under the signature ΣT Mt : {S = {S1}, E = {E1, E2, E3}, J = {F1, M1, M2, P2}, REJ = {E2, M1, E2, M2, E3, P2}, RJRE = {F1, S1, E1}}

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-24
SLIDE 24

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

First-order Relational Logic (FOL + Relational Calculus)

. Relational Join and ∼ Transpose The dot join and transpose operators ensure a uniform way of navigation between trace-locations through trace-links in constraints. ∗ˆ(Refmexive) Transitive Closure Transitive Closure allows the encoding of common reachability constraints that otherwise could not be expressed in FOL, such as preventing cyclic dependencies between trace-locations. Domain and Range Restrictions The restriction operators are used to fjlter relations to a given domain or range.

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-25
SLIDE 25

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

First-order Relational Logic (FOL + Relational Calculus)

. Relational Join and ∼ Transpose E.REJ = {E1, E2, E3}.{E2, M1, E2, M2, E3, P2} = {M1, M2, P2}

  • J. ∼ REJ = {F1, M1, M2, P2}.{M1, E2, M2, E2, P2, E3}

= {E2, E3} ∗ˆ(Refmexive) Transitive Closure ˆ{M1, E1, E1, C1} = {M1, E1, E1, C1, M1, C1} Domain and Range Restrictions P <: RJE = {P2} <: {M1, E2, M2, E2, P2, E3} = {P2, E3}

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-26
SLIDE 26

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-27
SLIDE 27

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Basic Type and SubType

⊥ ⊤ Requirement Implementation Code Executable ExecutableObjectCode ContractReq SystemReq none = { } univ = {r1,r2,r3,r4,i1,i2,i3}

denotes the set of all possible atoms denotes empty set and is the type of no atoms Top level signatures (basic type) Extension signatures (subtype)

in in

Subset signatures

LowLevelReq {r1}

Subtype (also subtype polymorphism or inclusion polymorphism) defines a subtyping relation, it is reflexive (meaning A<:A for any type A) and transitive (meaning that if A<:B and B<:C then A<:C). This makes it a preorder

  • n types.

{r1,r2} {r3} {r1,r2,r3,r4} {i1} {i1,i2}

in in

{i1} {i1,i2,i3}

The domain of discourse of any structure of that signature is then fragmented into disjoint subsets, one for every sort.

extends extends extends

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-28
SLIDE 28

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Relation Types

⊥ ⊤ Requirement Implementation Code Executable ExecutableObjectCode ContractReq SystemReq none = { } univ = {r1,r2,r3,r4,i1,i2,i3}

denotes the set of all possible atoms denotes empty set and is the type of no atoms

in in

LowLevelReq {r1} {r1,r2}

contain: ContractReq lone -> ContractReq

{r1,r2,r3,r4} {i1} {i1,i2}

in in

{i1} {i1,i2,i3} {r3}

satisfiedBy: SystemReq -> some Implementation refine: ContractReq one -> some SystemReq

{(r3,r1),(r3,r2)}

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-29
SLIDE 29

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-30
SLIDE 30

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Formal Specifjcation of an example confjguration

1 abstract sig Artefact { depends: set Artefact} 2 3

  • - Locate@File

4

  • ne sig Specification extends Artefact {

5 contract: some ContractRequirement} 6 7

  • - Locate@Text

8 sig ContractRequirement extends Artefact { 9 system: set SystemRequirement , 10 contains: set ContractRequirement} 11 12

  • - Locate@ReqIF

13 sig SystemRequirement extends Artefact { 14 satisfiedBy: set Implementation , 15 requires: set SystemRequirement , 16 refines: set SystemRequirement}

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-31
SLIDE 31

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

17 abstract sig Implementation extends Artefact { 18 fulfills: lone ContractRequirement} 19 20

  • - Locate@Java

21 sig Code, Component extends Implementation {} 22 23

  • - Locate@EMF

24 sig Model extends Implementation { 25 transforms , conforms: set Model, 26 generates: set (Code ∪ Component)} 27 28

  • - Semantics@SystemRequirement.satisfiedBy

29 fact {∀ i: Implementation | some i.˜satisfiedBy}

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-32
SLIDE 32

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Automated analysis functions over Traceability Model

Consistency Checking The system checks whether the user model satisfjes the specifjcation or not. Reasoning about Trace-relations If the model is a partial (incomplete), the platform tries to complete the model with respect to the semantics declared in the specifjcation inferring new trace-relations on the model. Trace-elements Discovery If a de-synchronization occurs on one or more ends of a trace-link probably caused by a change such as deletion of a trace-location, we try to repair the broken link based on the specifjed semantics.

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-33
SLIDE 33

Introduction Approach Demonstration Conclusion and Future Work Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

Reasoning about Trace-relations

30

  • - Reason@ContractRequirement.system

31 fact {∀ s: SystemRequirement , s': s.*˜refines | 32 s'.˜system = s.˜system} 33 34

  • - Reason@SystemRequirement.requires

35 fact { ∀ s, s': SystemRequirement | 36 s' in s.refines = ⇒ s in s'.requires } 37 38

  • - Reason@Implementation.fulfills

39 fact {∀ i: Implementation , s: i.˜satisfiedBy 40 | i.fulfills = s.˜system }

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-34
SLIDE 34

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-35
SLIDE 35

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Confjguration of User’s Workspace

User Traceability Framework Tarski Platform Alloy Configuration of Eclipse Workspace Formal Specification of the semantics alloy4compiler User’s Workspace Traceability Management Adaptation of Tarski Platform to Traceability Domain First-order Model Management Abstract Syntax Tree Automated Analysis of Traceability Automated Analysis Decision Procedure Interpretation Function Alloy Parser

Alloy Specification

Eclipse Front-End Analyze Custom Annotations Functions Consistency Check Reason about Relations Trace Elements Discovery

uses

Synthesis Internal Representation KodKod API Call Type hierarchy (sigs, fields) & Semantics (facts) KodKod Relational Model Finder Functions

interprets

Analyzing Traceability

Reasoning about trace instance

Load/Update Interface First-order Relational Model Relation Interpreted Atom Universe Tuple * * * Traceability Information *

* 1 from to elements

* Creating Trace Location

...

Text Fragments EClasses, EObjects XML Elements Java Elements Update/Delete Trace Location Assigning Types Loading/Updating Spec. Trace Location Trace Link Traceability *

The user can assign types to both trace locations and links using the relation names of the specification generates uses uses

Model

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-36
SLIDE 36

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Confjguration of User’s Workspace

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-37
SLIDE 37

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Type Hierarchy from the Specifjcation

User Traceability Framework Tarski Platform Alloy Configuration of Eclipse Workspace Formal Specification of the semantics alloy4compiler User’s Workspace Traceability Management Adaptation of Tarski Platform to Traceability Domain First-order Model Management Abstract Syntax Tree Automated Analysis of Traceability Automated Analysis Decision Procedure Interpretation Function Alloy Parser

Alloy Specification

Eclipse Front-End Analyze Custom Annotations Functions Consistency Check Reason about Relations Trace Elements Discovery

uses

Synthesis Internal Representation KodKod API Call Type hierarchy (sigs, fields) & Semantics (facts) KodKod Relational Model Finder Functions

interprets

Analyzing Traceability

Reasoning about trace instance

Load/Update Interface First-order Relational Model Relation Interpreted Atom Universe Tuple * * * Traceability Information *

* 1 from to elements

* Creating Trace Location

...

Text Fragments EClasses, EObjects XML Elements Java Elements Update/Delete Trace Location Assigning Types Loading/Updating Spec. Trace Location Trace Link Traceability *

The user can assign types to both trace locations and links using the relation names of the specification generates uses uses

Model

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-38
SLIDE 38

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Type Hierarchy from the Specifjcation

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-39
SLIDE 39

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-40
SLIDE 40

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Creating Trace-locations and Assigning Types

User Traceability Framework Tarski Platform Alloy Configuration of Eclipse Workspace Formal Specification of the semantics alloy4compiler User’s Workspace Traceability Management Adaptation of Tarski Platform to Traceability Domain First-order Model Management Abstract Syntax Tree Automated Analysis of Traceability Automated Analysis Decision Procedure Interpretation Function Alloy Parser

Alloy Specification

Eclipse Front-End Analyze Custom Annotations Functions Consistency Check Reason about Relations Trace Elements Discovery

uses

Synthesis Internal Representation KodKod API Call Type hierarchy (sigs, fields) & Semantics (facts) KodKod Relational Model Finder Functions

interprets

Analyzing Traceability

Reasoning about trace instance

Load/Update Interface First-order Relational Model Relation Interpreted Atom Universe Tuple * * * Traceability Information *

* 1 from to elements

* Creating Trace Location

...

Text Fragments EClasses, EObjects XML Elements Java Elements Update/Delete Trace Location Assigning Types Loading/Updating Spec. Trace Location Trace Link Traceability *

The user can assign types to both trace locations and links using the relation names of the specification generates uses uses

Model

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-41
SLIDE 41

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Assigning a Sub Type to a Trace-location

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-42
SLIDE 42

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Assigning a binary Field Type to a Trace-link

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-43
SLIDE 43

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Selecting a Trace-Location from the co-domain of the type

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-44
SLIDE 44

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Traceability Information

...

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-45
SLIDE 45

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-46
SLIDE 46

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

First-order Relational Model

...

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-47
SLIDE 47

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Dynamic Confjguration & Model Management

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-48
SLIDE 48

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Outline

1

Introduction Motivation Industrial Use Cases

2

Approach Traceability Domain Model First-order Relational Model and Logic Type Annotation and Trace-Relations Formal Semantics and Automated Analysis

3

Demonstration Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

4

Conclusion and Future Work

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-49
SLIDE 49

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Reasoning about Trace-instance

...

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-50
SLIDE 50

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Automated Analysis of Traceability

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-51
SLIDE 51

Introduction Approach Demonstration Conclusion and Future Work Formal Specifjcation of Traceability Semantics Traceability Management First-order Model Management Automated Analysis of Traceability

Synthesis of Internal Representation

...

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-52
SLIDE 52

Introduction Approach Demonstration Conclusion and Future Work

Should we consider also the temporal behavior of the traceability? Interesting analysis scenarios exist in industry We are not supporting ordered sets of Alloy which usually help model the dynamic behaviour. First-order theory of relations might be a candidate for traceability in Multi-pardigm Modeling for Cyber-physical

  • Systems. Preliminary results shows that the approach works
  • n the synchronization of design rules with design/installation
  • f physical components.

However, DPLL(T) solvers does not currently exists for this fragment of the theory. Alloy Language is too expressive for the domain of traceability. We’re working on the formalization of a First-order theory for traceability and the development of a domain-specifjc language for traceability.

  • F. Erata et al.

Traceability Analysis on Tarski Platform

slide-53
SLIDE 53

Introduction Approach Demonstration Conclusion and Future Work

Modeling and Reasoning Approaches

  • F. Erata et al.

Traceability Analysis on Tarski Platform