Fault Trees from Data: Efficient Learning with an Evolutionary Algorithm

Alexis Linard*, Doina Bucur, Mariëlle Stoelinga

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

Cyber-physical systems come with increasingly complex architectures and failure modes, which complicates the task of obtaining accurate system reliability models. At the same time, with the emergence of the (industrial) Internet-of-Things, systems are more and more often being monitored via advanced sensor systems. These sensors produce large amounts of data about the components’ failure behaviour, and can, therefore, be fruitfully exploited to learn reliability models automatically. This paper presents an effective algorithm for learning a prominent class of reliability models, namely fault trees, from observational data. Our algorithm is evolutionary in nature; i.e., is an iterative, population-based, randomized search method among fault-tree structures that are increasingly more consistent with the observational data. We have evaluated our method on a large number of case studies, both on synthetic data, and industrial data. Our experiments show that our algorithm outperforms other methods and provides near-optimal results.
Original languageEnglish
Title of host publicationDependable Software Engineering. Theories, Tools, and Applications
Subtitle of host publication5th International Symposium, SETTA 2019, Shanghai, China, November 27-29, 2019, Proceedings
EditorsNan Guan, Joost-Pieter Katoen, Jun Sun
Place of PublicationCham
PublisherSpringer
Pages19-37
Number of pages19
ISBN (Electronic)978-3-030-35540-1
ISBN (Print)978-3-030-35539-5
DOIs
Publication statusPublished - 2019
Event5th International Symposium on Dependable Software Engineering. Theories, Tools, and Applications, SETTA 2019 - Shanghai, China
Duration: 27 Nov 201929 Nov 2019
Conference number: 5

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume11951
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349
NameProgramming and Software Engineering
PublisherSpringer

Conference

Conference5th International Symposium on Dependable Software Engineering. Theories, Tools, and Applications, SETTA 2019
Abbreviated titleSETTA
CountryChina
CityShanghai
Period27/11/1929/11/19

Fingerprint

Fault Tree
Evolutionary algorithms
Evolutionary Algorithms
Sensors
Internet of Things
Sensor
Failure modes
Failure Mode
System Reliability
Synthetic Data
Tree Structure
Search Methods
Model
Learning
Experiments
Experiment

Keywords

  • Fault tree induction
  • Safety-critical systems
  • Cyber-physical systems
  • Evolutionary algorithm

Cite this

Linard, A., Bucur, D., & Stoelinga, M. (2019). Fault Trees from Data: Efficient Learning with an Evolutionary Algorithm. In N. Guan, J-P. Katoen, & J. Sun (Eds.), Dependable Software Engineering. Theories, Tools, and Applications: 5th International Symposium, SETTA 2019, Shanghai, China, November 27-29, 2019, Proceedings (pp. 19-37). (Lecture Notes in Computer Science; Vol. 11951), (Programming and Software Engineering). Cham: Springer. https://doi.org/10.1007/978-3-030-35540-1_2
Linard, Alexis ; Bucur, Doina ; Stoelinga, Mariëlle. / Fault Trees from Data : Efficient Learning with an Evolutionary Algorithm. Dependable Software Engineering. Theories, Tools, and Applications: 5th International Symposium, SETTA 2019, Shanghai, China, November 27-29, 2019, Proceedings. editor / Nan Guan ; Joost-Pieter Katoen ; Jun Sun. Cham : Springer, 2019. pp. 19-37 (Lecture Notes in Computer Science). (Programming and Software Engineering).
@inproceedings{cd1e1bd742874c8da5bbad02da596fa2,
title = "Fault Trees from Data: Efficient Learning with an Evolutionary Algorithm",
abstract = "Cyber-physical systems come with increasingly complex architectures and failure modes, which complicates the task of obtaining accurate system reliability models. At the same time, with the emergence of the (industrial) Internet-of-Things, systems are more and more often being monitored via advanced sensor systems. These sensors produce large amounts of data about the components’ failure behaviour, and can, therefore, be fruitfully exploited to learn reliability models automatically. This paper presents an effective algorithm for learning a prominent class of reliability models, namely fault trees, from observational data. Our algorithm is evolutionary in nature; i.e., is an iterative, population-based, randomized search method among fault-tree structures that are increasingly more consistent with the observational data. We have evaluated our method on a large number of case studies, both on synthetic data, and industrial data. Our experiments show that our algorithm outperforms other methods and provides near-optimal results.",
keywords = "Fault tree induction, Safety-critical systems, Cyber-physical systems, Evolutionary algorithm",
author = "Alexis Linard and Doina Bucur and Mari{\"e}lle Stoelinga",
year = "2019",
doi = "10.1007/978-3-030-35540-1_2",
language = "English",
isbn = "978-3-030-35539-5",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "19--37",
editor = "Nan Guan and Joost-Pieter Katoen and Jun Sun",
booktitle = "Dependable Software Engineering. Theories, Tools, and Applications",

}

Linard, A, Bucur, D & Stoelinga, M 2019, Fault Trees from Data: Efficient Learning with an Evolutionary Algorithm. in N Guan, J-P Katoen & J Sun (eds), Dependable Software Engineering. Theories, Tools, and Applications: 5th International Symposium, SETTA 2019, Shanghai, China, November 27-29, 2019, Proceedings. Lecture Notes in Computer Science, vol. 11951, Programming and Software Engineering, Springer, Cham, pp. 19-37, 5th International Symposium on Dependable Software Engineering. Theories, Tools, and Applications, SETTA 2019, Shanghai, China, 27/11/19. https://doi.org/10.1007/978-3-030-35540-1_2

Fault Trees from Data : Efficient Learning with an Evolutionary Algorithm. / Linard, Alexis; Bucur, Doina; Stoelinga, Mariëlle.

Dependable Software Engineering. Theories, Tools, and Applications: 5th International Symposium, SETTA 2019, Shanghai, China, November 27-29, 2019, Proceedings. ed. / Nan Guan; Joost-Pieter Katoen; Jun Sun. Cham : Springer, 2019. p. 19-37 (Lecture Notes in Computer Science; Vol. 11951), (Programming and Software Engineering).

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Fault Trees from Data

T2 - Efficient Learning with an Evolutionary Algorithm

AU - Linard, Alexis

AU - Bucur, Doina

AU - Stoelinga, Mariëlle

PY - 2019

Y1 - 2019

N2 - Cyber-physical systems come with increasingly complex architectures and failure modes, which complicates the task of obtaining accurate system reliability models. At the same time, with the emergence of the (industrial) Internet-of-Things, systems are more and more often being monitored via advanced sensor systems. These sensors produce large amounts of data about the components’ failure behaviour, and can, therefore, be fruitfully exploited to learn reliability models automatically. This paper presents an effective algorithm for learning a prominent class of reliability models, namely fault trees, from observational data. Our algorithm is evolutionary in nature; i.e., is an iterative, population-based, randomized search method among fault-tree structures that are increasingly more consistent with the observational data. We have evaluated our method on a large number of case studies, both on synthetic data, and industrial data. Our experiments show that our algorithm outperforms other methods and provides near-optimal results.

AB - Cyber-physical systems come with increasingly complex architectures and failure modes, which complicates the task of obtaining accurate system reliability models. At the same time, with the emergence of the (industrial) Internet-of-Things, systems are more and more often being monitored via advanced sensor systems. These sensors produce large amounts of data about the components’ failure behaviour, and can, therefore, be fruitfully exploited to learn reliability models automatically. This paper presents an effective algorithm for learning a prominent class of reliability models, namely fault trees, from observational data. Our algorithm is evolutionary in nature; i.e., is an iterative, population-based, randomized search method among fault-tree structures that are increasingly more consistent with the observational data. We have evaluated our method on a large number of case studies, both on synthetic data, and industrial data. Our experiments show that our algorithm outperforms other methods and provides near-optimal results.

KW - Fault tree induction

KW - Safety-critical systems

KW - Cyber-physical systems

KW - Evolutionary algorithm

UR - http://www.scopus.com/inward/record.url?scp=85076704235&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-35540-1_2

DO - 10.1007/978-3-030-35540-1_2

M3 - Conference contribution

SN - 978-3-030-35539-5

T3 - Lecture Notes in Computer Science

SP - 19

EP - 37

BT - Dependable Software Engineering. Theories, Tools, and Applications

A2 - Guan, Nan

A2 - Katoen, Joost-Pieter

A2 - Sun, Jun

PB - Springer

CY - Cham

ER -

Linard A, Bucur D, Stoelinga M. Fault Trees from Data: Efficient Learning with an Evolutionary Algorithm. In Guan N, Katoen J-P, Sun J, editors, Dependable Software Engineering. Theories, Tools, and Applications: 5th International Symposium, SETTA 2019, Shanghai, China, November 27-29, 2019, Proceedings. Cham: Springer. 2019. p. 19-37. (Lecture Notes in Computer Science). (Programming and Software Engineering). https://doi.org/10.1007/978-3-030-35540-1_2