An active inference approach to on-line agent monitoring in safety-critical systems

Autores
Martinez, Ernesto Carlos; Avila, Luis Omar
Año de publicación
2015
Idioma
inglés
Tipo de recurso
artículo
Estado
versión publicada
Descripción
The current trend towards integrating software agents in safety–critical systems such as drones, autonomous cars and medical devices, which must operate in uncertain environments, gives rise to the need of on-line detection of an unexpected behavior. In this work, on-line monitoring is carried out by comparing environmental state transitions with prior beliefs descriptive of optimal behavior. The agent policy is computed analytically using linearly solvable Markov decision processes. Active inference using prior beliefs allows a monitor proactively rehearsing on-line future agent actions over a rolling horizon so as to generate expectations to discover surprising behaviors. A Bayesian surprise metric is proposed based on twin Gaussian processes to measure the difference between prior and posterior beliefs about state transitions in the agent environment. Using a sliding window of sampled data, beliefs are updated a posteriori by comparing a sequence of state transitions with the ones predicted using the optimal policy. An artificial pancreas for diabetic patients is used as a representative example.
Fil: Martinez, Ernesto Carlos. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Santa Fe. Instituto de Desarrollo y Diseño (i); Argentina
Fil: Avila, Luis Omar. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Santa Fe. Instituto de Desarrollo y Diseño (i); Argentina
Materia
Active Inference
Bayesian Surprise
On-Line Monitoring
Twin Gaussian Processes
Nivel de accesibilidad
acceso abierto
Condiciones de uso
https://creativecommons.org/licenses/by-nc-nd/2.5/ar/
Repositorio
CONICET Digital (CONICET)
Institución
Consejo Nacional de Investigaciones Científicas y Técnicas
OAI Identificador
oai:ri.conicet.gov.ar:11336/6890

id CONICETDig_6abb86a301631d08d9fbcfea13a25dcf
oai_identifier_str oai:ri.conicet.gov.ar:11336/6890
network_acronym_str CONICETDig
repository_id_str 3498
network_name_str CONICET Digital (CONICET)
spelling An active inference approach to on-line agent monitoring in safety-critical systemsMartinez, Ernesto CarlosAvila, Luis OmarActive InferenceBayesian SurpriseOn-Line MonitoringTwin Gaussian Processeshttps://purl.org/becyt/ford/1.2https://purl.org/becyt/ford/1The current trend towards integrating software agents in safety–critical systems such as drones, autonomous cars and medical devices, which must operate in uncertain environments, gives rise to the need of on-line detection of an unexpected behavior. In this work, on-line monitoring is carried out by comparing environmental state transitions with prior beliefs descriptive of optimal behavior. The agent policy is computed analytically using linearly solvable Markov decision processes. Active inference using prior beliefs allows a monitor proactively rehearsing on-line future agent actions over a rolling horizon so as to generate expectations to discover surprising behaviors. A Bayesian surprise metric is proposed based on twin Gaussian processes to measure the difference between prior and posterior beliefs about state transitions in the agent environment. Using a sliding window of sampled data, beliefs are updated a posteriori by comparing a sequence of state transitions with the ones predicted using the optimal policy. An artificial pancreas for diabetic patients is used as a representative example.Fil: Martinez, Ernesto Carlos. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Santa Fe. Instituto de Desarrollo y Diseño (i); ArgentinaFil: Avila, Luis Omar. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Santa Fe. Instituto de Desarrollo y Diseño (i); ArgentinaElsevier2015-08info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionhttp://purl.org/coar/resource_type/c_6501info:ar-repo/semantics/articuloapplication/pdfapplication/pdfhttp://hdl.handle.net/11336/6890Martinez, Ernesto Carlos; Avila, Luis Omar; An active inference approach to on-line agent monitoring in safety-critical systems; Elsevier; Advanced Engineering Informatics; 29; 4; 8-2015; 1083-10951474-0346enginfo:eu-repo/semantics/altIdentifier/url/http://www.sciencedirect.com/science/article/pii/S1474034615000749info:eu-repo/semantics/altIdentifier/doi/10.1016/j.aei.2015.07.008info:eu-repo/semantics/altIdentifier/doi/info:eu-repo/semantics/openAccesshttps://creativecommons.org/licenses/by-nc-nd/2.5/ar/reponame:CONICET Digital (CONICET)instname:Consejo Nacional de Investigaciones Científicas y Técnicas2025-09-29T10:17:34Zoai:ri.conicet.gov.ar:11336/6890instacron:CONICETInstitucionalhttp://ri.conicet.gov.ar/Organismo científico-tecnológicoNo correspondehttp://ri.conicet.gov.ar/oai/requestdasensio@conicet.gov.ar; lcarlino@conicet.gov.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:34982025-09-29 10:17:34.566CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicasfalse
dc.title.none.fl_str_mv An active inference approach to on-line agent monitoring in safety-critical systems
title An active inference approach to on-line agent monitoring in safety-critical systems
spellingShingle An active inference approach to on-line agent monitoring in safety-critical systems
Martinez, Ernesto Carlos
Active Inference
Bayesian Surprise
On-Line Monitoring
Twin Gaussian Processes
title_short An active inference approach to on-line agent monitoring in safety-critical systems
title_full An active inference approach to on-line agent monitoring in safety-critical systems
title_fullStr An active inference approach to on-line agent monitoring in safety-critical systems
title_full_unstemmed An active inference approach to on-line agent monitoring in safety-critical systems
title_sort An active inference approach to on-line agent monitoring in safety-critical systems
dc.creator.none.fl_str_mv Martinez, Ernesto Carlos
Avila, Luis Omar
author Martinez, Ernesto Carlos
author_facet Martinez, Ernesto Carlos
Avila, Luis Omar
author_role author
author2 Avila, Luis Omar
author2_role author
dc.subject.none.fl_str_mv Active Inference
Bayesian Surprise
On-Line Monitoring
Twin Gaussian Processes
topic Active Inference
Bayesian Surprise
On-Line Monitoring
Twin Gaussian Processes
purl_subject.fl_str_mv https://purl.org/becyt/ford/1.2
https://purl.org/becyt/ford/1
dc.description.none.fl_txt_mv The current trend towards integrating software agents in safety–critical systems such as drones, autonomous cars and medical devices, which must operate in uncertain environments, gives rise to the need of on-line detection of an unexpected behavior. In this work, on-line monitoring is carried out by comparing environmental state transitions with prior beliefs descriptive of optimal behavior. The agent policy is computed analytically using linearly solvable Markov decision processes. Active inference using prior beliefs allows a monitor proactively rehearsing on-line future agent actions over a rolling horizon so as to generate expectations to discover surprising behaviors. A Bayesian surprise metric is proposed based on twin Gaussian processes to measure the difference between prior and posterior beliefs about state transitions in the agent environment. Using a sliding window of sampled data, beliefs are updated a posteriori by comparing a sequence of state transitions with the ones predicted using the optimal policy. An artificial pancreas for diabetic patients is used as a representative example.
Fil: Martinez, Ernesto Carlos. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Santa Fe. Instituto de Desarrollo y Diseño (i); Argentina
Fil: Avila, Luis Omar. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Santa Fe. Instituto de Desarrollo y Diseño (i); Argentina
description The current trend towards integrating software agents in safety–critical systems such as drones, autonomous cars and medical devices, which must operate in uncertain environments, gives rise to the need of on-line detection of an unexpected behavior. In this work, on-line monitoring is carried out by comparing environmental state transitions with prior beliefs descriptive of optimal behavior. The agent policy is computed analytically using linearly solvable Markov decision processes. Active inference using prior beliefs allows a monitor proactively rehearsing on-line future agent actions over a rolling horizon so as to generate expectations to discover surprising behaviors. A Bayesian surprise metric is proposed based on twin Gaussian processes to measure the difference between prior and posterior beliefs about state transitions in the agent environment. Using a sliding window of sampled data, beliefs are updated a posteriori by comparing a sequence of state transitions with the ones predicted using the optimal policy. An artificial pancreas for diabetic patients is used as a representative example.
publishDate 2015
dc.date.none.fl_str_mv 2015-08
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
http://purl.org/coar/resource_type/c_6501
info:ar-repo/semantics/articulo
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/11336/6890
Martinez, Ernesto Carlos; Avila, Luis Omar; An active inference approach to on-line agent monitoring in safety-critical systems; Elsevier; Advanced Engineering Informatics; 29; 4; 8-2015; 1083-1095
1474-0346
url http://hdl.handle.net/11336/6890
identifier_str_mv Martinez, Ernesto Carlos; Avila, Luis Omar; An active inference approach to on-line agent monitoring in safety-critical systems; Elsevier; Advanced Engineering Informatics; 29; 4; 8-2015; 1083-1095
1474-0346
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/url/http://www.sciencedirect.com/science/article/pii/S1474034615000749
info:eu-repo/semantics/altIdentifier/doi/10.1016/j.aei.2015.07.008
info:eu-repo/semantics/altIdentifier/doi/
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
https://creativecommons.org/licenses/by-nc-nd/2.5/ar/
eu_rights_str_mv openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by-nc-nd/2.5/ar/
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.publisher.none.fl_str_mv Elsevier
publisher.none.fl_str_mv Elsevier
dc.source.none.fl_str_mv reponame:CONICET Digital (CONICET)
instname:Consejo Nacional de Investigaciones Científicas y Técnicas
reponame_str CONICET Digital (CONICET)
collection CONICET Digital (CONICET)
instname_str Consejo Nacional de Investigaciones Científicas y Técnicas
repository.name.fl_str_mv CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicas
repository.mail.fl_str_mv dasensio@conicet.gov.ar; lcarlino@conicet.gov.ar
_version_ 1844614129136959488
score 13.070432