logo EDITE Costin CAVAL
Costin CAVAL
État académique
Thèse soutenue le 2016-05-31
Sujet: Gestion des fautes imprévues dans les systèmes critiques
Direction de thèse:
Ellipse bleue: doctorant, ellipse jaune: docteur, rectangle vert: permanent, rectangle jaune: HDR. Trait vert: encadrant de thèse, trait bleu: directeur de thèse, pointillé: jury d'évaluation à mi-parcours ou jury de thèse.
Productions scientifiques
Keeping a Clear Separation between Goals and Plans
International audience
Many approaches to BDI agent modeling permit the agent developers to interweave the levels of plans and goals. This is possible through the adoption of new goals inside plans. These goals will have plans of their own, and the definition can extend on many levels. From a software development point of view, the resulting complexity can render the agents’ behavior difficult to trace, due to the combination of elements from different abstraction levels, i.e., actions and goal adoptions. This has a negative effect on the development process when designing and debugging agents. In this paper we propose a change of approach that aims to provide a more comprehensible agent model with benefits for the ease of engineering and the fault tolerance of agent systems. This is achieved by imposing a clear separation between the reasoning and the acting levels of the agent. The use of goal adoptions and actions on the environment inside the same plan is therefore forbidden. The approach is illustrated using two theoretical scenarios as well as an agent-based maritime patrol application. We argue that by constraining the agent model we gain in clarity and traceability therefore benefiting the development process and encouraging the adoption of agent-based techniques in industrial contexts.
Engineering Multi-Agent Systems https://hal.archives-ouvertes.fr/hal-01097697 Dalpiaz, Fabiano; Dix, Jürgen; van Riemsdijk, M.Birna. Engineering Multi-Agent Systems, 8758, Springer International Publishing, 2014, Lecture Notes in Computer Science, 978-3-319-14483-2. <10.1007/978-3-319-14484-9_2>Book sections 2014
A Multi-Agent Middleware for Deployment of Ambient Applications
International audience
This chapter overviews the actual use of connected devices and ambient applications. It proposes a multi-agent middleware to ease the development, deployment, configuration and monitoring of applications for ambient systems. To achieve Smart Cities, we need an operational and interoperable telecommunication and energetic infrastructure. We also need some intelligent and context-aware applications. The chapter reviews modeled the projection problem for the deployment of applications on a hardware infrastructure for ambient systems as an enriched graph-homomorphism problem. This allows us to reuse one of the existing and well-studied algorithms for graph-matching. The scenario used in the chapter highlights both the dynamic deployment of distributed applications and privacy management encapsulated in both the agents and agent organizations. To solve the dynamic deployment problem, one can use the graph representation for the hardware infrastructure and the requirements of the usable applications, and the graph matching algorithm to find the entities that can support the running of the application.
Enablers for Smart Cities https://hal.archives-ouvertes.fr/hal-01369635 Enablers for Smart Cities, 2016, 9781848219588. <10.1002/9781119329954.ch4>ARRAY(0x7f04009ec9b0) 2016-07
A Multi-Agent System for Resource Privacy: Deployment of Ambient Applications in Smart Environments
AAMAS - Extended Abstract
In this paper, we present Multi-Agent Systems (MAS) as a well-adapted paradigm for designing software for the configuration, deployment and monitoring of distributed applications in the domain of Ambient Intelligence (AmI). We show how privacy is enhanced by hiding information using the agent architecture and organisation. We introduce privacy policies that can allow or prevent the sharing of resource information. This results in privacy by design.
https://hal.archives-ouvertes.fr/hal-01369668 AAMAS - Extended Abstract. 2016ARRAY(0x7f0401b64520) 2016
A Multi-Agent Solution for the Deployment of Distributed Applications in Ambient Systems
International audience
Ambient Intelligence (AmI) and Internet of Things (IoT) arepromising fields for the application of Multi-Agent Systems (MAS). Aspecific MAS application, described through a video doorkeeper scenarioin this paper, is the deployment and the configuration of distributed ap-plications on a hardware infrastructure in ambient systems. It requiresthe modelling of the available infrastructure and of the deployable ap-plications, respecting a domain ontology, which can then be used byreasoning tools to find the hardware entities that can support the run-ning of the application in the existing infrastructure. It also requires adistributed architecture that allows this solution to be scalable and toprovide mechanisms to enhance privacy. In this paper, we discuss this lastpoint. We describe the use of goal-driven agents and show how the MASarchitecture and organisation allow for the privacy of the infrastructureresources to be enhanced.
Engineering Multi-Agent Systems https://hal.archives-ouvertes.fr/hal-01369674 Engineering Multi-Agent Systems, 2016ARRAY(0x7f0401b5be70) 2016
A Multi-Agent Approach for the Deployment of Distributed Applications in Smart Environments
IDC Conference paper
This paper presents an approach for the configuration, de-ployment and monitoring of distributed applications in a smart envi-ronment. This approach takes into consideration the heterogeneity andthe dynamicity of such environments and deals with resource privacy.We propose to describe the available hardware infrastructure and thedeployable applications using graphs, and provide a mathematical for-malisation of the deployment process based on graph homomorphisms. Adecentralised version of a branch and bound graph-matching algorithmis used to find the available hardware entities of the infrastructure thatcan be used to run the application, respecting its requirements. At last,we describe a goal-directed Multi-Agent System (MAS) for the deploy-ment of applications in ambient systems. We show that the multi-agentparadigm is well-adapted to provide a clear separation between the ap-plicative and the hardware layers, thus increasing resource privacy.
https://hal.archives-ouvertes.fr/hal-01369679 IDC Conference paper. 2016ARRAY(0x7f04025b29e8) 2016
Thèse: Agents Dirigés par des Buts pour la Tolérance aux Fautes Imprévues - Un Filet de Sécurité pour les Programmeurs
Soutenance: 2016-05-31
Rapporteurs: Adina Magda FLOREA    René MANDIAU