Now showing 1 - 10 of 17
  • Publication
    DevOps in Robotics: Challenges and Practices
    ( 2023)
    Sawczuk da Silva, Alexandre
    ;
    ; ;
    Rothe, Johannes
    ;
    Ihrke, Christoph
    DevOps, which refers to a set of practices for streamlining the development and operations of software companies, is becoming increasingly popular as businesses strive to adopt a loosely coupled architecture that supports frequent software delivery. As a result, DevOps is also gaining traction in other domains and involved architectures, including robotics, though research in this area is still lacking. To address this gap, this paper investigates how to adapt key DevOps principles from the domain of software engineering to the domain of robotics. In order to demonstrate the feasibility of this in practice, an industrial robotics case study is conducted. The results indicate that the adoption of these principles is also beneficial for robotic software architectures, though general DevOps approaches may require some adaptation to match the existing infrastructure.
  • Publication
    Towards Uncertainty Reduction Tactics for Behavior Adaptation
    An autonomous system must continuously adapt its behavior to its context in order to fulfill its goals in dynamic environments. Obtaining information about the context, however, often leads to partial knowledge, only, with a high degree of uncertainty. Enabling the systems to actively reduce these uncertainties at run-time by performing additional actions, such as changing a mobile robot’s position to improve the perception with additional perspectives, can increase the systems’ performance. However, incorporating these techniques by adapting behavior plans is not trivial as the potential benefit of such so-called tactics highly depends on the specific context. In this paper, we present an analysis of the performance improvement that can theoretically be achieved with uncertainty reduction tactics. Furthermore, we describe a modeling methodology based on probabilistic data types that makes it possible to estimate the suitability of a tactic in a situation. This methodology is the first step towards enabling autonomous systems to use uncertainty reduction in practice and to plan behavior with more optimal performance.
  • Publication
    Evaluation of Traffic Control Systems as ITS Infrastructure for Automated Driving
    ( 2018)
    Franze, Juliane
    ;
    Seydel, Dominique
    ;
    ;
    Haspel, Ulrich
    Vehicles with automated driving systems require more sensor information about their environment than non-automated vehicles. Detection with camera, lidar or other sensors is already state of the art in newer vehicles. As of today though, they only work in close proximity and lack the incorporation of existing traffic information from local authorities. In this paper, we present a novel way of providing traffic management information to vehicles, sent directly from Road Authorities. We use existing ITS (Intelligent Transport Systems) infrastructure and assess how information on traffic control and reroutes, displayed on variable message signs, can be used as sensory input for vehicles. We examine real world data from a South German Road Authority. The evaluation of latency, reliability and integrity of traffic information has been conducted end-to-end as well as between the six stations that are involved. We show the general feasibility of our proposal and discuss which obstacles need to be overcome for a wider use in other road systems.
  • Publication
    Safe adaptation for reliable and energy-efficient E/E architectures
    ( 2017) ; ; ;
    Ruiz, Alejandra
    ;
    Radermacher, Ansgar
    The upcoming changing mobility paradigms request more and more services and features to be included in future cars. Electric mobility and highly automated driving lead to new requirements and demands on vehicle information and communication (ICT) architectures. For example, in the case of highly automated driving, future drivers no longer need to monitor and control the vehicle all the time. This calls for new fault-tolerant approaches of automotive E/E architectures. In addition, the electrification of vehicles requires a flexible underlying E/E architecture which facilitates enhanced energy management. Within the EU-funded SafeAdapt project, a new E/E architecture for future vehicles has been developed in which adaptive systems ensure safe, reliable, and cost-effective mobility. The holistic approach provides the necessary foundation for future invehicle systems and its evaluation shows the great potential of such reliable and energy-efficient E/E architectures.
  • Publication
    Software implementieren und absichern
    Ein plötzliches Verkehrshindernis kann zur Gefahr werden. Vor allem dann, wenn es Autofahrer zu spät wahrnehmen und es nicht schaffen, rechtzeitig darauf zu reagieren. Ein Gefahrenwarner kann dem vorbeugen. Jedoch sind solche fahrzeugübergreifenden Funktionen sehr komplex: Sie erfordern neue Entwicklungsansätze und Entwicklungswerkzeuge. Das heißt vor allem, dass vernetzte Fahrfunktionen schon früh im Entwurf simuliert und getestet werden sollten.
  • Publication
    Ausfallsichere E/E-Architektur für hochautomatisierte Fahrfunktionen
    Die Hochautomatisierung erfordert neue Ansätze zur Ausfallsicherheit von Fahrzeugbordnetzen: Wird der Fahrer künftig von der Überwachung des Fahrzeugs befreit, müssen die E/E-Architekturen eine höhere Ausfallsicherheit bereitstellen. Im Rahmen des EU Projekts SafeAdapt wird hierzu ein ganzheitlicher Ansatz vorgestellt.
  • Publication
    Neue Konzepte für zukünftige E/E-Architekturen
    ( 2016) ;
    Rosenthal, Thorsten
    Zukünftige Fahrzeuge stehen aufgrund der sich rasant wandelnden Mobilität völlig neuen Herausforderungen gegenüber. Insbesondere die Anforderungen an die kommenden E/E-Architekturen verändern sich ständig. Im EU-geförderten Projekt SafeAdapt haben neun Partner aus sechs Ländern untersucht, inwiefern Adaptivität für flexible und zuverlässige E/E-Architekturen genutzt werden kann.
  • Publication
    Design of self-adaptation in distributed embedded systems
    (Verlag Dr. Hut, 2015) ;
    Knorr, Rudi
    ;
    Bauer, Bernhard
    Nowadays, complex computers are integrated in numerous devices and are deployed in diverse application areas, for example in the automotive, avionic, health-care, and industrial automation domain. These embedded systems are evolving towards complex interconnected and adaptive systems. They increasingly integrate more and more functionality and must function under varying conditions and in diverse situations. Therefore, distributed embedded systems become self-adaptive so that they can adjust to varying context situations. This poses new challenges for the development of these self-adaptive distributed embedded systems. Therefore, this thesis introduces a novel model-driven approach for designing self-adaptation of these systems. The presented approach is applied to the application domain of automotive Electrics / Electronics (E/E).
  • Publication
    Planetares Zusammenspiel
    ( 2014)
    Zeller, Marc
    ;
    Die zunehmende Anzahl an Funktionen in vernetzten eingebetteten Systemen erhöht auch die Komplexität der Software kontinuierlich. Die Modularisierung der Funktionen in einzelne Softwarekomponenten und standardisierte Softwarearchitekturen sollen dem entgegenwirken. Dabei ist es wichtig, die Softwareschnittstellen frühzeitig im Entwicklungsprozess zu definieren und abzusichern, um die Qualität und Fehlerfreiheit des Systems zu gewährleisten.
  • Publication
    Bessere Softwaremodelle mit einem Richtlinien-Katalog
    Mit dem von Fraunhofer ESK entwickelten Werkzeugprototypen können Entwickler die Einhaltung von Richtlinien bei der Modellierung von Softwaremodellen künftig verlässlicher, umfassender und schneller prüfen. Das Programm umfasst sowohl derzeit gängige Methoden für die Validierung von Softwaremodellen als auch Mechanismen für die Prüfung von Modellierungsrichtlinien: Ein modularer Richtlinienkatalog, Profile mit unterschiedlichen Regelanforderungen und die Konfiguration einzelner Regeln erlauben eine Anpassung an die aktuelle Modellierungsaufgabe. Dazu kommen Möglichkeiten zur Dokumentation und automatisierten Korrektur von gefundenen Problemen.