March 2025 - informatik-vortraege - lists.rwth-aachen.de

Einladung Informatik-Oberseminar Imke Nachmann
by Sek 18 Mar '25

18 Mar '25

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Donnerstag, 27.03.2025, 14:00 Uhr Ort: Informatikzentrum der RWTH Aachen University, Gebäude E3, 2. Etage, Raum 9222 Referentin: Imke Nachmann, M.Sc. Lehrstuhl für Software Engineering Thema: Functional Modeling of Cyber-Physical Systems Abstract: Engineering Cyber-Physical Systems faces many challenges including the demand to integrate many different forms of functionalities and features. Traditional engineering processes are structured by the physical components and software modules of the final system and the engineering activities are concerned with evolving these components. In these approaches, the information, which components realize which functions and, in particular, which components interact in which way to implement a function, most often remains implicit. This hinders not only the collaboration of experts from different domains but also an agile approach to engineering which becomes efficient through automation of tasks. The shift towards functional requirements gives rise to a conceptual gap between their abstract descriptions and the very detailed descriptions of the system’s implementation. So far, the involved engineering domains have established an understanding of what a system’s function is, but these understandings have not been consolidated and integrated in a suitable interdisciplinary modeling technique. Therefore, we propose a functional development paradigm that promotes development by finding realizations of cyber-physical functions, which transform energy, matter, and data, rather than implementing physical products or software modules directly. Therein, we model a as a network of interacting timed stream processing functions, which specify the desired logical behavior and the physical behavior. By applying this modeling technique to formalize the design process in mechanical engineering, we show how to integrate the functional understanding of software and mechanical systems, which paves the way for agile and holistic model-driven engineering of CPSs. Further, we propose requirements and a meta-model for implementing modeling languages to model these functions Es laden ein: die Dozentinnen und Dozenten der Informatik

1 0

Einladung / Informatik-Oberseminar / Rohit Gupta (Dissertationsvortrag)
by rg.gupta＠siemens.com 14 Mar '25

14 Mar '25

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Donnerstag, 10. April 2025, 14 Uhr Ort: Room 9222, Building E3, Ahornstr. 55 Referent: Rohit Gupta M.Sc. Chair of Software Engineering (Computer Science 3) Thema: A Systematic Approach to Fostering Engineering of Industrial Domain-Specific Modelling Languages Abstract: Domain-Specific Modelling Languages (DSMLs) help modellers and domain experts in various domains such as healthcare, energy, information technology, and so on, in reducing the gap between the problem space and the solution space by placing models at the centre of development activities. This shift towards model-driven development (MDD), where models are introduced at early stages in any software or systems engineering process and are the primary software engineering artefacts, allows modellers to design complex, heterogeneous real-world abstractions of their systems. In such complex systems, concepts from individual domains are often integrated as part of the bigger language infrastructure. Specifically in an industrial setting, the methodologies to describe a systematic engineering process for developing such complex yet modular and reusable DSMLs that provides a seamless modelling experience to modellers in both the large scale organisations as well as in small and medium enterprises is still largely neglected. Accordingly, this thesis is aimed at providing the means to engineer graphical DSMLs that are specifically focussed on industrial contexts. Based on existing approaches, this work presents a systematic approach to fostering the engineering of industrial DSMLs by composing reusable language infrastructure parts without the need for creating completely new language infrastructure for similar domains every time. These reusable units of a DSML, termed DSML building blocks, consist of reusable language components that, entirely or in part, contributes to the technical definition of the language itself. The language components compose through different forms of language composition to form heterogeneous, integrated DSMLs. To foster the interoperability of such common language infrastructure parts between modelling environments, a bidirectional exchange mechanism is detailed in this work. This work further provides guidelines and design decisions that language engineers should consider for their language infrastructure in order to elevate the overall experience of modellers. An approach for integrating methods, techniques, and concepts in terms of guidance and recommendations for modellers is detailed in this thesis that aims to move away from plain technical views of models to instead model-aware and dynamic views that are focussed on the current modelling situation of such modellers. Overall, this thesis presents approaches to modularly build reusable units of DSMLs that compose together. The approaches presented in this thesis allows language engineers to provide a more complete and integrated language infrastructure that is ultimately aimed at improving the modelling experience of practitioners in the industry. Es laden ein: die Dozentinnen und Dozenten der Informatik

1 0

Einladung: Informatik-Oberseminar Christian Martin Fiedler
by DSME-Office 12 Mar '25

12 Mar '25

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Dienstag, 25. März 2025, 14.00 Uhr Ort: Seminarraum 423, 4. OG, Theaterstraße 35-39, 52062 Aachen Referent: Christian Martin Fiedler M.Sc. Institut für Data Science im Maschinenbau (DSME) Thema: Contributions to Kernel Methods in Systems and Control Abstract: Machine learning is increasingly used in systems and control, which is motivated by challenging control, simulation and analysis problems, abundant data and computing resources, as well as impressive theoretical and methodological advances in machine learning. The established class of kernel methods is of particular interest in this context, due to their rich theory, modularity, efficient and reliable algorithms, and indeed kernel methods are frequently used in systems and control. We present two exemplary and complementary contributions to this flourishing field. First, many learning-based control approaches are based on combining uncertainty bounds for Gaussian process (GP) regression with robust control methods. We revisit the foundations of this domain by consolidating, improving, and carefully evaluating the required uncertainty bounds, and use them in learning-enhanced control applications. Furthermore, we discuss a severe practical limitation of these approaches, the a priori knowledge of an upper bound on the reproducing kernel Hilbert space (RKHS) norm of the target function, and propose to combine geometric assumptions together with kernel methods as a promising alternative. Second, we initiate a new research direction by combining kernels with mean field limits as appearing in kinetic theory. Motivated by learning problems on large-scale multiagent systems, we introduce mean field limits of kernels, and provide an extensive theory for the resulting RKHSs. This is used in turn in the analysis of kernel-based statistical learning in the mean field limit, which not only is a novel form of large-scale limit in theoretical machine learning, but provides also a solid foundation for applications in kinetic theory. Es laden ein: die Dozentinnen und Dozenten der Informatik

1 0