Logo des Repositoriums

Auflistung nach Autor:in "Sudeikat, Jan"

1 - 3 von 3
  • Konferenzbeitrag
    Controlled Run-Time Adaptivity in Industrial Agent Systems - Challenges and Research Prospects
    (INFORMATIK 2023 - Designing Futures: Zukünfte gestalten, 2023) Sudeikat, Jan; Köhler-Bußmeier, Michael
    Developing Cyber-physical Systems (CPS) inherently requires enabling run-time adaptivity. These systems integrate physical components, which operate in changeable contexts. In addition, objectives may change, due to socio-technical aspects. Industrial agents have been proposed for enhancing future industrial automation and control systems and integrating agents in industrial systems is an active field of research. Besides enabling technical compatibility of agent concepts and frameworks, the design and orchestration of agent activities have to be fine-tuned and a goal-directed adaption@run.time requires that the system can analyze its own structure at run-time; therefore, the system’s structure has to be reflected inside the system. Here, we outline current work and research challenges on how explicit organizational modeling can facilitate developing industrial agent systems. We discuss architectural aspects and outline how adaptations of organizations can be enabled, modeled and automated following the MAPE-K approach.
  • Konferenzbeitrag
    Toward the Design of Self–Organizing Dynamics
    (Software Engineering 2007 – Beiträge zu den Workshops – Fachtagung des GI-Fachbereichs Softwaretechnik, 2007) Sudeikat, Jan
    A growing demand for distributed and decentralized software, together with an increasing inherent complexity of these systems challenges traditional software engineering practices. Recently, the utilization of self–organizing processes has been proposed as a way to relieve development efforts. While Agent–Oriented Software Engineering (AOSE) provides suitable abstractions and implementation frameworks for these system dynamics, the design of phenomena that rise solely from entity interactions contradicts traditional top–down development strategies. This paper describes work in progress and outlines a doctoral thesis that addresses the analysis and design of self–organizing distributed systems in Multi–Agent Systems (MAS), i. e. collections of autonomous, pro–active entities.
  • Konferenzbeitrag
    Towards simulation-supported enterprise architecture management
    (Modellierung betrieblicher Informationssysteme (MobIS 2008), 2008) Buckl, Sabine; Matthes, Florian; Renz, Wolfgang; Schweda, Christian M.; Sudeikat, Jan
    Enterprise architecture management is based on a holistic view on the enterprise addressing business and IT aspects in an integrated manner. EA management is a process to manage the complexity of the overall architecture and to improve the alignment of business and IT. In order to achieve these goals, it is necessary but not sufficient to manage the static complexity that arises from dependencies between the elements of the EA, like goals, organizational units, business processes, business applications, and IT infrastructure elements. Performance, stability, and scalability can only be analyzed, modeled, and controlled, if static EA models are enriched by appropriate abstractions to capture the dynamic complexity of the EA understood as a socio-technical system of interacting (sub-)systems. This article identifies possible techniques to address dynamic complexity in EA. The potential benefits of system simulations are discussed and the derivation of apropriate simulation models is exemplified. A key observation is the fact that EA management is a iterative evolution process, where each iteration only changes a small fraction of the EA. It is therefore possible to automatically derive model parameters required for the simulation of the future architectures from an analysis of the dynamics of the current architecture.