Zum Hauptinhalt springen

Mitarbeiter*innenverzeichnis

Stefan Bosse

Forschungsprofil

Forschungserfahrung und Projektleitung im Bereich:

  • Verteilte Künstliche Intelligenz Übertragung von KI und biologisch inspirierten Konzepten auf technische Systeme mit parallelen und verteilten Eigenschaften unter besonderer Beachtung von Interaktion/Kommunikation und Skalierbarkeit (Effizienz, Komplexität).
  • Pervasive & Ubiquitous Computing, Internet der Dinge, Strukturüberwachung (SHM) Crowd- und Things Sensing spielt eine immer größere Rolle in der globalen Informationsgewinnung, und zunehmend mit Zugriff und Verwertung von ohnehin erfassten Sensordaten.
  • Maschinelles Lernen & Data Mining Vor allem inkrementelle und verteilte Lernverfahren für unzuverlässige Daten (z.B. Sensordaten) stellen ein zentrales Forschungsthema dar um skalierbare und effiziente Lernverfahren in verteilten Systemen einsetzen zu können, die lokale Perzeption und globales Schließen durch Kooperation erlauben.  Weiterhin ist Methodenfusion ein wichtiges Thema (Numerik+ML+MAS).
  • Multiagenten- und selbstorganiserende Systeme Mobile Multiagentensysteme werden für robuste und adaptive Informationsverarbeitung in stark heterogenen Umgebungen eingesetzt.
  • Simulation Simulation von großskaligen/komplexen technischen und soziologischen Systemen (Sozioinformatik), vor allem mittels agentenbasierter Simulation und Simulation von Multiagentensystemen, kombiniert mit physikalischer Simulation (Mehrbereichssimulation).
  • Material Informatik Material Informatik aus einer anderen Sicht die Berechnung in Materialien und technische Strukturen bringen soll um zukünftige Intelligente Materialien und Strukturen entwickeln und einsetzen zu können.
  • Sensornetzwerke, Signalverarbeitung, Sensor, Materialintegration Alle technischen Ebenen der verteilten Sensorverarbeitung, Sensortechnologien, Kommunikationstechnologien und Konzepte

Weitere Kompetenzen

  • Modellbasierter Entwurf eingebetteter Systeme, Rechnerarchitektur, Schaltkreisentwurf Einsatz von HDL-basierten Entwurfsmethoden und Highlevel Syntheseverfahren (HLS) für den Entwurf komplexer SoC
  • Cyber Physical Systems, Robotische Systeme Anwendung und technische Realisierung der verteilten KI in robotischen Systemen in Produktion, Logistik, Haushalt, usw.
  • Programmiersprachen und Compilerbau (HLS) Entwurf von Programmiersprachen und Entwicklung von Synthesewerkzeugen
  • Materialwissenschaften Materialintegration von Datenverarbeitungsnetzwerken (z.B. für SHM)
  • Numerische Verfahren

Lehre

Studiengänge Informatik, Systems Engineering, Produktionstechnik

  • Universtät Bremen: Parallele und verteilte eingebettete Systeme, Entwurf eingebetteter Systeme, Multiagentensysteme
  • Universität Koblenz: Grundlagen der Funktionalen Programmierung (Grundstudium), Verteilte und parallele Programmierung, Multiagentensysteme

Sensorische Materialien

  • Entwicklung von Algorithmen und Methoden zur kontextabhängigen Informationsgewinnung aus Sensordaten.
  • Entwicklung von Kommunikationsstrategien in Sensornetzwerken und Ableitung von Kommunikationsprotokollen.
  • Simulation und Bewertung von Kommunikationsprozessen in Sensornetzwerken.
  • Zusammenfassende Betrachtung und Simulation verbundener Energiequellen, Verbraucher und Speicher in Sensornetzwerken.
  • Entwicklung und Prüfung (software- und hardware-in-the-loop) von Energiemanagement-Konzepten für Drahtlose und drahtgebundene Informations- und Energieübertragung.

Konferenzen und Zeitschriften

  • Organisation von Konferenzen,
  • Gast Editor in internationalen Zeitschriften

Entwurf von Programmiersprachen

  • ConPro: Concurrent Programming, Parallele Programmiersprache für den Digital Hardware Entwurf von SoC unter Benutzung von High-level Syntheseverfahren
  • SEM: SeSAm Simulation Language, Textuelle Repräsentation von Verhaltensmodellen für Multi-Agenten Systemen und dem SeSAm Simulator
  • AAPL: Activity-Transition Graph based Agent Programming Language, generische Programmiersprache für die Modellierung von mobilen zustandsbasierten Multi-Agenten Systemen, welche in heterogenen Netzwerken eingesetzt werden können
  • AFL: Agent FORTH, stackorientierte FORTH Programmiersprache, basierend auf AAPL,
  • AgentJS: Agent JavaScript, basierend auf<

    Research

    Research experience and project leadership in the fields:

    • Distributed AI
    • Multi-agent Systems and their technological deployment
    • Self-organizing and self-adaptive Systems
    • Agent Platforms
    • Machine Learning
    • Distributed and Parallel Systems
    • Simulation of and with Multi-agent Systems
    • Development and design of Sensorial adn Self-adaptive Materials; Material Science
    • Sensor Networks with embedded systems
    • Design of parallel and distributed embedded systems
    • Digital circuit design with high-level synthesis approaches
    • Sensor signal processing
    • Virtual Machines
    • Compiler Design

    Lectures

    Lectures in bachelor and master courses teaching basic and higher-level knowledge of computer science, circuit design, and distributed and parallel systems, multi-agent systems and distributed AI.

    • University of Bremen: Parallel and distributed embedded systems, Design of embedded systems, Multi-agent Systems
    • University of Koblenz: Introduction to Functional Programming, Distributed and parallel Programming, Multi-agent Systems

    Conferences and Journals

    • Organization of international conferences, i.e., SysInt 2014 Conference held in Bremen, ECSA-2, Basel (November 2015)
    • Guest Editor in several international journals, i.e., IEEE Sensors, Elsevier Mechattronics

    Design of Programming Langauges

    • ConPro: Concurrent Programming, Parallel programming language for digital hardware designs and SoC using high-level synthesis
    • SEM: SeSAm Simulation Language, Textual representation of behaviorual models for multi-agent systems and the SeSAm simulator
    • AAPL: Activity-based Agent Programming Language, generic programming language for modelling of mobile state-based multi-agent systems, which are deplyoed in heterogeneous networks
    • AFL: Agent FORTH, a stack-based FORTH programming language with agent behaviour related to the ATG model, migration, and tuple space interaction
    • AgentJS: Agent JavaScript, based on AAPL, programming language for the JAM agent platform
    • VPL: Virtual Database Programming Language, Interface programming language for the virtual graph database VDB that is deployed in the Synthesis Toolkit SynDK used for the construction of complex compiler and synthesis frameworks.
    • JavaScript Semantic Type System (JST/JS+)

    Software

Stefan Bosse studied physics at the University of Bremen. He received a PhD/doctoral degree (Dr. rer. nat.) in physics in the year 2002 (topic "Advanced Optical Laser Measuring Techniques") at the University of Bremen, and the post-doctoral degree (Habilitation) and the Venia Legendi in Computer Science in the year 2016 at the University of Bremen with his habilitation (postdoctoral degree) "Unified Distributed Sensor and Environmental Information Processing with Multi-Agent Systems: Models, Platforms, and Technological Aspects".

Since 2017 he is teaching and researching as a Privatdozent at the University of Bremen, Department of Computer Science, and since 2018 he is an interim professor at the University of Koblenz-Landau, Faculty Computer Science, Institute of Software Technologies.

At the University of Bremen and University Koblenz he teaches several courses in fundamental computer science, functional programming, and in selected advanced topics covering the design and programming of massive parallel and distributed systems, multi-agents systems and agent-based simulation, high-level synthesis of complex digital logic data processing systems, and material-integrated sensing systems with a high interdisciplinary background.

His main research area is distributed artificial intelligence in general, and in particular information processing in massive parallel and distributed systems using agent-based approaches combined with machine learning, and agent-based simulation. A broad range of fields of application and domains are addressed: Material Science, Materials Informatics, Smart Materials, IoT, Production Engineering, Social Science, Crowd Sensing, Geo Science.

He conducted projects in the internationally recognized ISIS Scientific Centre for Intelligent Sensorial Materials pushing interdisciplinary research closing the gap between technology and computer science, finally joining the ISIS council and publishing an internationally well regarded handbook on this topic.

He acts as a reviewer and a guest editor for several international journals and is a member of a broad range of international conference programme and organizing committees.

Stefan Bosse studied physics at the University of Bremen. He received a PhD/doctoral degree (Dr. rer. nat.) in physics in the year 2002 (topic "Advanced Optical Laser Measuring Techniques") at the University of Bremen, and the post-doctoral degree (Habilitation) and the Venia Legendi in Computer Science in the year 2016 at the University of Bremen with his habilitation (postdoctoral degree) "Unified Distributed Sensor and Environmental Information Processing with Multi-Agent Systems: Models, Platforms, and Technological Aspects".

Since 2017 he is teaching and researching as a Privatdozent at the University of Bremen, Department of Computer Science, and since 2018 he is an interim professor at the University of Koblenz-Landau, Faculty Computer Science, Institute of Software Technologies.

At the University of Bremen and University Koblenz he teaches several courses in fundamental computer science, functional programming, and in selected advanced topics covering the design and programming of massive parallel and distributed systems, multi-agents systems and agent-based simulation, high-level synthesis of complex digital logic data processing systems, and material-integrated sensing systems with a high interdisciplinary background.

His main research area is distributed artificial intelligence in general, and in particular information processing in massive parallel and distributed systems using agent-based approaches combined with machine learning, and agent-based simulation. A broad range of fields of application and domains are addressed: Material Science, Materials Informatics, Smart Materials, IoT, Production Engineering, Social Science, Crowd Sensing, Geo Science.

He conducted projects in the internationally recognized ISIS Scientific Centre for Intelligent Sensorial Materials pushing interdisciplinary research closing the gap between technology and computer science, finally joining the ISIS council and publishing an internationally well regarded handbook on this topic.

He acts as a reviewer and a guest editor for several international journals and is a member of a broad range of international conference programme and organizing committees.