January 27, 2016
Technical Talk at Meeting of Working Group Leaders of Silicon Saxony
Fog Computing is a new reference architecture for sensor networks at the edge of the cloud. Its basic idea is to process data locally, in complex sensor nodes, gateways and routers, before moving them into the cloud. Thereby, it protects privacy by default, and delivers speed (low latency, high bandwidth). Its business perspectives for Germany are tremendous, because starting from the sensor markets, Germany companies can try to penetrate into the future cloud markets. And combined with LED-based transmission of data (LiFi), it will also result in highly energy-efficient infrastructures for Industry 4.0, smart homes and smart environments. Watch out!
Nov 03, 2014
Technical Talk at the seminar of the Collaborative Research Center HAEC
So far, autotuning has been a form of continuous optimization for specific kernels and algorithms. In the Collaborative Research Center "Highly-Adaptive Energy-Efficient Computing (HAEC)", we develop a generalized form of autotuning for software product lines. The approach is based on cost-utility functions and relations, which are specified in quality contracts. A more specific form treats energy-utility functions, which describe constraints on energy behavior. From these contracts, constraint-based systems are generated to be solved by constraint solvers. Generalized autotuning attributes every variant of a software product line with quality and energy contracts, and then decides at run time of the application, which variant is the most appropriate with regard to a specific objective function. This generalizes autotuning from the level of specific kernels to dynamic software product lines.
Oct 28, 2014
RoSI PhD training group, Kickoff meeting
Roles compartmentalize objects by context.
This compartmentalization helps both on the object and the type level. The number of aliases are reduced, the lifetime of attributes is better known, and the state of an object naturally decomposes by context. Roles therefore improve modeling and programming in many ways.
The RoSI PhD training group investigates many research questions for role-based languages and infrastructures.
May 05, 2012
Link÷pings Universitet, Sweden
We are going from fly-by-wire to drive-by-wire to life-by-wire. Many aspects of our life are already controlled by software and electronics, and many more will be in the future. In this talk, we investigate the technical requirements for reliable cyber-physical systems in the future internet of things (iot). We show that CPS must be self-adaptive to changing requirements, while nevertheless offering full reliability and safety. This can be mastered with MOO architectures based on multi-objective optimization. We also look at the market mechanisms and software platforms for life-by-wire and the resulting software ecosystems. A new global player is searched for the platform leadership for cyber-physical systems.
December 17, 2009
Technical Talk at Workshop at University of Twente, The Netherlands
We discuss an aspect-oriented decomposition scheme for software, Essence-Administration-Infrastructure (EAI), from Steve McMenamin. We show its similarity to Composition Filters and show how to employ it for simple evolution of software systems.
Feb 13, 2009
Technical Talk at Queens University, Kinston, Canada
Jakob Henriksson, Jendrik Johannes, Steffen Zschaler and Uwe A▀mann
Software languages differ in their support for software reuse - some offer pretty poor mechanisms (C, Cobol,...) and others are very smart (BETA, Scala,UML,...). Wouldn't it be good to be able to separate the reuse mechanism from the core language? Can we define something like Reuse Languages (RL) that can be combined with algorithmic or declarative core languages to support universal reuse mechanisms?
Universal invasive software composition is a new technology to specify reuse languages. Module systems, class systems, fragment systems and other reuse techniques can be developed for languages in a universal way. Given a grammar or metamodel of a core language, a component model can be specified, from which a composition system can be generated that offers sophisticated forms of reuse of all software artefacts written in the core language. Since this works language-universally, all languages, even if they do not offer reuse mechanisms, can be equipped with reuse technology: reuse for the reuse-agnostic. Several examples are presented based on the Reuseware system (www.reuseware.org).
Technical Talk at Dagstuhl Seminar "XML and Ownership Types"
Prof. Dr. Uwe A▀mann http://www.rewerse.net
For the future Semantic Web, an integration of ontologies into standard languages is urgendly needed. This talk presents a concept for the integration of ontologies as domain models into the MDA process (Model-Driven Architectures). In this way, they can form the basis of a product line.
The talk is supported by the EU 6th framework Network of Excellence REWERSE http://www.rewerse.net
Technical Talk at Lin÷pings Universitet
Dr. Uwe A▀mann, Martin Nilsson, Leif Stensson, Marcus Comstedt http://www.gotpike.org https://github.com/pikelang http://pike.lysator.liu.se/
For the future Semantic Web, languages for rapid application development are urgendly needed. This talk presents a concept for the integration of ontologies into the scripting language Pike, which is hosted at Link÷pings Universitet http://www.gotpike.org
Dr. Uwe A▀mann, Prof. Dr. Peter Fritzson
The first-generation web appeared in 1990 and brought an industrial revolution - all document formats in all industries have changed since then. We claim that the next industrial revolution will be provoced by the "Semantic Web", an initiative started by the W3C and supported by the European Commission. The Semantic Web technology adds typing to the documents of the future and will serve for better interoperability and type-checking of documents and specifications in all industries. To be early on the train, we propose a Swedish Semantic Web initiative.