EPiCS is implemented by a consortium of 8 institutions from 5 countries:
- University of Paderborn
- Imperial College, London
- University of Oslo
- Klagenfurt University
- University of Birmingham
- EADS Innovation Works Munich
- ETH Zurich
- AIT Austrian Institute of Technology
University of Paderborn, Germany
University of Paderborn (UPB) will be represented within the project by the Paderborn Center for Parallel Computing (PC²). PC² was founded in 1991 as an interdisciplinary institute of UPB and since then has established a reputation as a competence centre for parallel and distributed computing and innovative computer architecture. A specific strength of PC² is its focus on innovative computer architecture research, in particular in custom and reconfigurable computing. The personnel involved in EPiCS has an extraordinary expertise in these areas, with more than 100 publications at major international conferences and journals in the field.
University of Paderborn's role in the consortium is:
- Project coordination and management.
- Development of an architecture and operating system for a self-aware computing platform.
- Contribution to the development of the self-aware network architecture.
Prof. Marco Platzner (Project Coordinator)
Dr. Christian Plessl (EPiCS scientific and technical manager)
Imperial College, London, United Kingdom
Imperial College London is one of the leading research universities in Europe specialised in engineering, medicine, natural sciences and business. The Department of Computing at Imperial is well known for work on reconfigurable technology and custom computing, high-performance computing, distributed computing, logic and artificial intelligence, visual information processing, and computing theory. Professor Wayne Luk leads the Custom Computing Group which focuses on design methods, languages, tools and models for custom hardware, parallel computers and embedded systems, particularly those involving reconfigurable devices such as field-programmable gate arrays, and their applications in finance, communication, and medical imaging.
Imperial's role in the EPiCS project is to develop methods and tools for self-optimising and self-verifying systems that can extract and tune application parameters, in order to perform dynamic optimisations in self-aware systems. These techniques will contribute to hardware and software platforms for such systems in EPiCS. A heterogeneous compute cluster will be used as a demonstrator to illustrate our approach.
Prof. Wayne Luk
University of Oslo, Norway
University of Oslo (UIO) participates in the EPiCS project through the Robotics and Intelligent Systems (ROBIN) research group at the Department of Informatics and the Department of Musicology. Our competence is within the area of adaptable hardware systems, bio-inspired computing, machine learning and musicology explored in by the fourMs lab.
Our role in the consortium is to:
- Develop self-learning, self-organizing and adaptable systems through the application of bio-inspired computing and machine learning techniques. Exploiting self-awareness to anticipate changes in the environment is also included.
- Develop a mobile media system providing an active music experience (called hypermusic) based on the listener’s motion and interaction with neighboring units.
EPiCS website at UIO
Prof. Jim Tørresen
Klagenfurt University, Austria
The Pervasive Computing Group is part of the Institute of Networked and Embedded Systems at Klagenfurt University. As the name of the group implies, the work and research is focused on pervasive computing. This includes topics like digital signal processors and RFIDs as well as sensor networks, smart cameras and sensor fusion.
For the EPiCS-project, the Klagenfurt University will create a distributed smart camera system in collaboration with the Austrian Institute of Technology (AIT) to detect and track objects autonomously within the network. This system will use a novel tracking algorithm, developed by the AIT, to identify and track objects within the field of view of a camera. The Klagenfurt University will implement a middleware to incorporate self-awareness concerning resources, tracked objects and available cameras. Tracing an object within a distributed camera network, where the object is only tracked by one camera at the time, is one of the biggest challenges in the project. If the object leaves the field of view of the first camera and enters the field of view of another camera, the second camera has to track flawlessly. To accomplish that, the identification of spatial relations between fields of view is most important. This should happen dynamically and during runtime to allow cameras to join and leave the network without needing any further human interaction.
Prof. Bernhard Rinner (EPiCS dissemination and exploitation manager)
University of Birmingham, United Kingdom
The University of Birmingham was the UK’s first 'redbrick' university, and is a member of the Russell Group of leading UK research-intensive universities. Representing the University of Birmingham is Cercia, a world-class research centre specialising in evolutionary computation and neural network ensemble learning. It carries out both fundamental research and knowledge transfer to industrial, commercial and public sector partners. Its academic research has been recognised internationally by prestigious best paper awards and some of the most cited papers in computer science. Cercia’s expertise lies in the development and study of computational systems that use ideas and get inspiration from the natural world, including biological, ecological, economical and physical systems. Research strengths include evolutionary optimisation, co-evolutionary learning, ensemble learning, evolutionary neural networks, evolvable hardware, self-organisation and emergent behaviour, self-adaptive systems, and search-based software engineering.
Within EPiCS, Cercia’s primary role is to bring its expertise to bear in studying the concepts and foundations of future self-aware and self-expressive systems, in particular:
- online learning in dynamic, self-organising environments;
- mechanisms and algorithms to ensure that large decentralised self-organising systems give rise to desirable global behaviours through local interactions only;
- the effect of interacting nodes’ objectives, strategies and behaviour on overall robustness, performance and quality of service in a complex and decentralised network;
- methods for exploiting self-awareness in a complex adaptive system in order to learn to anticipate changes in a dynamic environment.
Prof. Xin Yao
EADS Innovation Works Munich, Germany
EADS Innovation Works is the corporate research center of the European Aeronautic Defence and Space Company. Its mission is to research, develop and integrate emerging technologies for transfer into innovative products of the EADS core business units Airbus, Eurocopter, Astrium, and Cassidian. EADS IW research topics cover a broad range of scientific disciplines from material sciences and production processes through electronics, simulation and IT services.
In the context of the EPiCS project, EADS IW's contributions are centered around reconfigurable HW/SW platforms, associated operating system services and programming models. A heavy emphasis lies on the application of proprioceptive and reconfigurable systems in safety-critical systems, in particular in aircraft electronics (avionics). Thus, EADS IW is investigating fault-tolerance concepts such as thread-level redundancy, checkpointing and rollback of hard- and software threads, and fault-tolerant communication mechanisms that rely on the developed self-awareness and self-expression concepts.
Dr. Stephan Stilkerich (EPiCS scientific and technical manager)
ETH Zurich, Switzerland
ETH Zurich has come to symbolise excellent education, groundbreaking basic research and applied results that are beneficial for society as a whole. Founded in 1855, ETH Zurich has more than 15,000 students from approximately 80 countries, 3,400 of whom are doctoral candidates. Today, it offers researchers an inspiring environment and students a comprehensive education as one of the world’s leading universities for technology and the natural sciences. 21 Nobel Laureates have studied, taught or conducted research at ETH Zurich, underscoring the excellent reputation of the institute.
In the EPiCS project ETH Zurich is represented by the Communication Systems Group (CSG). The CSG is one of three research groups of the Computer Engineering and Networks Lab (TIK), which itself is the largest lab of the Department of Information Technology and Electrical Engineering (D-ITET).
The CSG is participating in the EPiCS project in order to continue its research in autonomic networking. The focus lies in the introduction of self-awareness and self-expression in computer networks and how this can be optimally realized in hardware.
Prof. Bernhard Plattner
AIT Austrian Institute of Technology – Safety & Security Department, Austria
The AIT Austrian Institute of Technology is Austria's largest non-university research institute. Among the European research institutes, AIT is a specialist in the key infrastructure issues of the future. The Safety & Security Department makes a significant contribution to ICT and devotes concerted efforts to guaranteeing operational efficiency and reliability of all critical infrastructures.
The group's research and development is focused on (embedded) software and hardware for video and security applications. This encompasses the development of camera and encoder platforms, the development of interactive tools for video exploration as well as research on innovative algorithms for intelligent video monitoring.
The work for this project tackles significant and unsolved Computer Vision problems - the visual representation, detection and tracking of a human and the camera calibration in large-scale camera networks. We focus on novel approaches that on the one hand incorporate the self-awareness or rather on-line learning paradigm and on the other hand link detection, tracking and calibration. The ultimate goal is to develop interactive tools for infrastructure operators enabling effective video surveillance.
Dr. Roman Pflugfelder