Models@run.time: A Guided Tour of the State of the Art and Research - - PowerPoint PPT Presentation

▶
models run time a guided tour of the state of the
SMART_READER_LITE
LIVE PREVIEW

Models@run.time: A Guided Tour of the State of the Art and Research - - PowerPoint PPT Presentation

Dec 21, 2022 30 likes •385 views

Models@run.time: A Guided Tour of the State of the Art and Research Challenges* Motivation: Problems and Objectives Research Problems A plethora of approaches has been presented since the term Models@run.time was coined in 2006 Work

slide-1
SLIDE 1

Models@run.time: A Guided Tour of the State of the Art and Research Challenges*

slide-2
SLIDE 2

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 2

Motivation: Problems and Objectives Research Problems

  • A plethora of approaches has been presented since the term

„Models@run.time“ was coined in 2006

  • Work on Models@run.time is distributed among many venues
  • By now, only one overview article from 2013 on the research topic exists

Research Objectives

  • Thus, a new overview article is required, which
  • Guides researchers new to the field
  • Helps the researchers to position their work
slide-3
SLIDE 3

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 3

Contributions of the Study Based on these objectives, our study offers the following contributions:

  • 1. A novel taxonomy to classify work on models@run.time
  • 2. A quantitative overview of the current state-of-the-art
  • 3. A catalogue of open research challenges

based on gaps in the state of the art

slide-4
SLIDE 4

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 4

Research Method

slide-5
SLIDE 5

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 5

Research Method

  • An adjusted version of Kitchenham‘s Systematic Literature Review:
  • Venue-based search to get initital literature corpus
  • Initial taxonomy based on experience
  • In-/exclusion criteria-based filtering
  • Evolving the taxonomy while iterating the literature corpus
  • adding classes to dimensions
  • Splitting/merging dimensions
  • Classifying the literature alongside
  • Keword-based search to include papers from further venues
  • Cross-dimensional gap-analysis
slide-6
SLIDE 6

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 6

In-/Exclusion Criteria Inclusion Criteria

  • The paper covers research where a model, which reflects the state of a

system, should be causally connected with that system.

  • The paper addresses runtime models or explicitly uses the term

models@run.time.

  • The paper uses self-representation, reflection or self-modelling.

Exclusion Criteria

  • Approaches on executable models are not to be considered

models@run.time approaches, if they lack the causal connection to the system, but are the actual system.

  • The survey includes papers published until December 2017.
slide-7
SLIDE 7

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 7

Overview of Literature Corpus

Name Type #AllPapers #Included

Models@run.time Workshop 95 80 Requirements@run.time Workshop 13 11 RAM-SE Workshop 61 6 MRT Dagstuhl Book 11 9 MODELS Conference 545 23 SEAMS Conference 182 29 ICAC Conference 366 31 SASO Conference 385 11 CompArch Conference 330 7 ECSA/WICSA Conference 334 8 RE Conference 528 6 SPLC Conference 120 2 ICSE Conference 640 6 MRT Special Issues Journal 25 9 SoSyM Journal n/a 3 JSS Journal n/a 2 TOSEM Journal n/a 1 TSE Journal n/a 4 TAAS Journal n/a 4

GoogleScholar Search

n/a n/a 20 Total 3635 272

slide-8
SLIDE 8

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 8

Overview of Included Papers

  • Notably, most papers are published at conferences
  • Since 2008 until 2017, every year at least 15 papers on models@run.time

have been published

slide-9
SLIDE 9

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 9

The Taxonomy

slide-10
SLIDE 10

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 10

A Novel Taxonomy for Models@run.time

Taxonomy of Research on Models@run.time Type of Research Applied Research Modelled Artifacts Types of Used Models Purpose of Runtime Models Applied MDSD Techniques Fundamental Research Domain of Application Intersecting Research Areas Related Research Initiatives

slide-11
SLIDE 11

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 11

The Taxonomy – Modelled Artifacts

slide-12
SLIDE 12

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 12

The Taxonomy – Types of Runtime Models

slide-13
SLIDE 13

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 13

The Taxonomy – Purposes of Using Runtime Models

slide-14
SLIDE 14

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 14

The Taxonomy – Applied Model-driven Techniques

slide-15
SLIDE 15

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 15

The Taxonomy – Fundamental Research Topics

slide-16
SLIDE 16

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 16

The Taxonomy – Application Domains, Research Areas and Initiatives Application Domains

  • Most work (23) has been evaluated in an enterprise software context
  • In total we found 25 different application domains used for evaluation

Intersecting Research Areas

  • We found 25 intersecting research areas, confirming the hypothesis that

models@run.time is highly interdisciplinary

  • The most prominent research area is „self-adaptive systems“ (78)

Initiatives

  • We found a very large number of research projects (19 EU, 26 regional)
  • But, only few projects directly focus on models@run.time
slide-17
SLIDE 17

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 17

Analysis

slide-18
SLIDE 18

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 18

Cross-dimensional Gap-Analysis

  • To identify research gaps, we performed a cross-dimensional gap-analysis
  • We compared the four dimensions of applied research on models@run.time

Dimension #1 Dimension #2

slide-19
SLIDE 19

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 19

12 13 96 10 13 feedback loop (1) metamodel (0) goal (14) design (6) requirement (6) behavior (45) variability (10) physical (4) structure (128) quality (23) component (26) process (12) code (16) context (20) gui (1) architecture (130) goals (32) Type of Runtime Model Modelled Artifact

Analysis – Modelled Artifact vs. Type of Model

[OC1] [OC2] [OC3] [OC4] [OC5] [OC11]

slide-20
SLIDE 20

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 20

Analysis – Modelled Artifact vs. Purpose of Runtime Model

17 20 61 11 25 assurance (38) development (28) fault-tolerance (7) prediction (6) self- adaptation (113) interoperability (7) evolution (15) self-optimization and organization (17) component (26) process (10) code (16) context (20) gui (1) architecture (126) goals (32) Purpose of Runtime Model Modelled Artifact [OC11] [OC3]

slide-21
SLIDE 21

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 21

25 30 18 16 10 14 10 reflection (41) model- transformation (48) workflow (4) reasoning (32) variability- modelling (22) machine- learning (14) monitoring (22) analysis (30) reqeng (10) model- comparison (4) component (26) process (11) code (16) context (18) gui (1) architecture (123) goals (32) Applied Model-driven Technique Modelled Artifact

Analysis – Modelled Artifact vs. Applied Model-driven Technique

[OC11] [OC10] [OC9]

slide-22
SLIDE 22

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 22

13 14 20 15 18 56 11 12 10

assurance (18) development (11) fault-tolerance (0) prediction (0) self- adaptation (51) interoperability (2) evolution (5) self-optimization and organization (3) feedback loop (6) metamodel (3) goal (14) design (7) requirement (6) behavior (44) variability (10) physical (4) structure (125) quality (23) Purpose of Runtime Model Type of Runtime Model

Analysis – Model Type vs. Purpose of Runtime Model

[OC6] [OC7]

slide-23
SLIDE 23

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 23

12 27 35 19 13 12

reflection (42) model- transformation (50) workflow (6) reasoning (32) variability- modelling (22) machine- learning (14) monitoring (23) analysis (30) requirements enginerring (10) model- comparison (4) feedback loop (1) metamodel (3) goal (14) design (7) requirement (4) behavior (44) variability (10) physical (3) structure (126) quality (21) Applied Model-driven Technique Type of Runtime Model

Analysis – Model Type vs. Applied Model-driven Technique

[OC10] [OC5] [OC9]

slide-24
SLIDE 24

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 24

17 11 20 21 14 17

assurance (38) development (26) fault-tolerance (7) prediction (6) self- adaptation (110) inter-

  • perability (9)

evolution (17) self-optimization and organization (17) requirements enginerring (10) model- comparison (4) model- comparison (4) reflection (41) model- transformation (50) workflow (5) reasoning (32) variability- modelling (22) machine- learning (13) monitoring (23) analysis (30) requirements engineering (10) Purpose of Runtime Model Applied Model-driven Techjnique

Analysis – Purpose of Runtime Model vs. Applied Model-driven Technique

[OC9] [OC11] [OC10]

slide-25
SLIDE 25

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 25

Research Challenges

slide-26
SLIDE 26

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 26

Future Research Challenges – Artifact-based

  • The need to apply models@run.time at lower levels than architecture
  • Especially, on code-level (e.g., working with the AST) [OC1]
  • The need to apply models@run.time at higher levels than architecture [OC2]
  • Systems-of-systems
  • Collective self-aware systems
slide-27
SLIDE 27

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 27

Future Research Challenges – Model-Type-based

  • Goal models at runtime to address uncertainty [OC3]
  • Variability models at runtime (SPLC and DSPL workshop series) [OC4]
  • Runtime feedback loop models [OC5]
slide-28
SLIDE 28

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 28

Future Research Challenges – Based on Purposes

  • Assurance, especially for safety-critical systems (i.e., certification) [OC6]
  • Development (seamless integration with development-time models) [OC7]
  • Self-aware computing systems (especially, collectives thereof) [OC8]
slide-29
SLIDE 29

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 29

Future Research Challenges – Based on Techniques

  • Machine Learning [OC9]
  • Systematic Model Comparison [OC10]
  • Runtime Workflow Models (trend stopped 5 years ago) [OC11]
slide-30
SLIDE 30

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 30

Future Research Challenges – Fundamental Research

  • Towards Managed Uncertainty at Runtime
  • Runtime Model Inference
  • Inferring the runtime model from the running system
  • Runtime Code Synthesis
  • Inferring the running system from the runtime model
  • Distributed Models@run.time
  • Transaction-safe Causal Connections
  • Self-modelling systems
slide-31
SLIDE 31

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 31

Overview of Future Directions

Level of Abstraction Type of Runtime Model Purpose of Runtime Model Applied Model-driven Technique Code-level Models Runtime Context Models Runtime Models for Systems-of-Systems Goal Models Variability Models Feedback Loops Providing Assurances Machine Learning Model Comparison Workflow Modeling Fundamental Research Topics Runtime Model Inference Managed Uncertainty Distributed Runtime Models Runtime Code Synthesis Self-modelling Systems Transaction-safe Causal Connections Using Development Models at Runtime Self-awareness

slide-32
SLIDE 32

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 32

Tool Support

slide-33
SLIDE 33

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 33

The Systematic Literature Review Toolkit

  • To automate common manual tasks in our study, we developed an Eclipse-

based RCP: the SLR Toolkit

  • Tool Demo on Friday, 10:00, 2A56 (Aud 3)
  • Tool paper in this years MODELS Proceedings
slide-34
SLIDE 34

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 34

slide-35
SLIDE 35

Models@run.time: A Guided Tour of the State-of-the-Art and Research Challenges Nelly Bencomo, Sebastian Götz, Hui Song 13th International Workshop on Models@run.time @ MODELS‘18, Copenhagen, Denmark // 14.10.2018 Slide 35

Thank you!