Technical Reports

Citation

(2002), "Technical Reports", Kybernetes, Vol. 31 No. 1. https://doi.org/10.1108/k.2002.06731aab.011

Technical Reports

Technical Reports

EPSRC E-Science Pilot Projects (Received 9 OCT. 01)

As part of the Spending Review Settlement, SR 2000, the United Kingdom Government announced a major investment of some £120M for a new initiative in e-Science. This initiative is concerned with the development of key IT and computing infrastructure to support the increasingly global research collaborations that are emerging in many areas of science and engineering. Such collaborations are based on the shared use of some combination of very large computing resources, enormous data collections and remote use to specialized facilities or sensor data. The need for such experiments to access extreme computing resources and/or multi-petabyte datasets, together with their associated visualization requirements, will in turn drive the development of the next generation of computing infrastructure. Such infrastructure will also benefit areas of business, commerce and education.

The UKs Engineering and Physical Sciences Research Council (EPSRC) wiII manage, on behalf of all the Research Councils, a core e-Science Programme, the purpose of which is to identify and develop generic grid middleware, in collaboration with industry. In addition, EPSRC – alongside the other Research Councils – received £17M for the development of multi-disciplinary grid application projects aimed at testing grid technology in a number of distinct science and engineering areas, in collaboration with industry and commerce. Following a call for proposals, EPSRC has allocated this £17M to six successful projects. Companies are contributing a further £8M in cash and in-kind to these projects. The successful projects are:

Distributed Aircraft Maintenance Environment (DAME)

Principal Investigator: Professor Jim Austin, Dept of Computer Science, University of York. Email: Austin@cs.york.ac.uk

Collaborating Universities: Oxford (Dept of Electrical Engineering), Sheffield (Automatic Control and Systems Engineering), Leeds (School of Mechanical Engineering and Dept of Computing).

Collaborating Companies: Rolls Royce plc; Data Systems and Solutions. Further companies are expected to join the project.

The project aims to build a Grid test-bed for distributed diagnostics. The application demonstrator will initially be a distributed aircraft maintenance environment for the maintenance of civil aircraft engines. The technology is, however, also applicable to other sectors such as medical, pharmaceutical, transport and manufacturing. The project will capitalize on experience of data grids in the USA and will address performance issues such as large-scale dam management. The essential research theme of the work is real-time intelligent feature extraction, intelligent data mining and decision support tools which are distributed on a global basis. The scale of the data generated by the 10,000 aircraft engines in flight, the size of the data bases needed to handle the information and the need for distributed access to these data make it a particularly demanding challenge for grid technologies. The Universities already collaborate through the White Rose Computational Grid currently under construction at a cost of £2.8M to those universities.

Structure-Property Mapping: Combinatorial Chemistry and the Grid

Principal Investigator: Dr Jeremy Frey, Dept of Chemistry, University of Southampton. Email: J.G.Frey@soton.ac.uk

Collaborating Universities: Southampton (Dept of Electronics & Computer Science and Mathematical Studies) Bristol (Chemistry).

Collaborating Companies: Roche Discovery Welwyn; Pfizer; IBM UK Ltd; Cambridge Crystallographic Data Centre; Southampton Combinatorial Centre of Excellence.

The goal of the project is to develop an e-Science test-bed that integrates existing structure and property data sources, and augments them within a grid-based information and knowledge environment. The synthesis of new chemical compounds by combinatorial methods provides major opportunities for the generation of large volumes of new chemical knowledge and is the principal drive behind the project. An extensive range of primary data needs to be accumulated, integrated and relationships modeled, so that maximum knowledge can be derived. The service-based grid computing infrastructure extends to devices in the laboratory and involves enriched systems, (including multimedia and live metadata), full support for provenance and innovative techniques for automation throughout the environment. The results of the project will impact on the design of materials through the prediction of properties and the identification of suitable compounds in a variety of applications.

The Reality Grid – A Tool for Investigating Condensed Matter and Materials

Principal Investigator. Professor Peter Coveney, Dept of Chemistry, Queen Mary and Westfield College, London. Email: P.V.Coveney@qmw.ac.uk

Collaborating Universities: Edinburgh (Dept of Physics and Astronomy), Manchester Computing Centre, Imperial College, Loughborough (Dept of Computer Science), London (Dept of Computer Science), Oxford (Dept of Materials), Manchester (Dept of Computer Science).

Collaborating Companies: CfS Consortium; Schlumberger Cambridge Research Ltd; Edward Jenner Institute for Vaccine Research; Silicon Graphics Inc; Advanced Visual Systems Ltd; FECIT. Collaborating Universities: Southampton (Dept of Electronics & Computer Science and Mathematical Studies) Bristol (Chemistry).

The project aims to enable the realistic modeling of complex condensed matter structures at the meso and nano-scale levels and for the discovery of new materials. The project has immediate extensions into e-Bioscience and e- Climate. The Reality Grid extends the concept of a virtual reality center across the grid and links it to massive computational resources at HPC centers and experimental facilities. Using grid technology to closely couple high throughput experimentation and visualization, the Reality Grid will move the bottleneck out of the hardware and back to the human mind. A two-track approach will be used. A ''fast track'' to use currently available grid middleware to construct a working grid; and a ''deep track'' which involves computer science teams in harnessing leading-edge research to create a robust and flexible problem-solving environment in which to embed the Reality Grid.

Grid Enabled Optimisation and Design Search for Engineering (GEODISE)

Principal Investigator: Dr Stephen Cox, Dept of Electronics and Computer Science, University of Southampton. Email: sc@ecs.soton.ac.uk

Collaborating Universities: Oxford (Computing Laboratory), Manchester (Dept of Computer Science).

Collaborating Companies: Rolls Royce plc; BAE Systems plc; Fluent Europe Ltd; Intel Corp (UK) Ltd; Microsoft Ltd; Epistemics Ltd; Compusys plc.

Engineering design search and optimization is the process whereby engineering modeling and analysis are exploited to yield improved designs. In the next 2-5 years, intelligent search tools will become a vital component of all engineering design systems and will steer the user through the process of setting up, executing and post-processing design search and optimization activities. Whilst the evaluation of a single design may require the analysis of gigabytes of data, to improve the process of design can require assimilation of terabytes of distributed data. Achieving the latter will lead to the development of intelligent search tools. GEODISE will provide grid-based seamless access to an intelligent knowledge repository, a state of the art collection of optimization and search tools, industrial strength and analysis codes, and distributed computing and data resources. This will involve large-scale distributed simulations to be coupled with tools to describe and modify designs, often in real time, using information from a knowledge base. Such tools will be physically distributed and under control of multiple elements in the design supply chain. The initial industrial application for the project will be on the use of computational fluid dynamics.

Discovery Net: An E-Science Test-Bed for High Throughput Informatics

Principal Investigator: Dr Yike Guo, Dept of Computing, Imperial College of Science, Technology and Medicine, London. Email: yg@doc.ic.ac.uk

Collaborating Depts within Imperial College: Bioengineering, Physics, Biochemistry, Electrical and Electronic Engineering, Huxley School of Environment, Earth Sciences and Engineering. Collaborating Companies: Inforsense Ltd; Deltadot Ltd; RVCO Inc.

The Discovery Net project aims to design, develop, and implement an advanced computing infrastructure to support real-time processing, interpretation, integration, visualization and mining of massive amounts of time-critical data generated by high throughput devices. The project will encompass new technology devices including biochips in biology, high throughput screening technology in biochemistry and combinatorial chemistry, high throughput sensors in energy and environmental science, remote sensing and geology. Application areas will include, analysis of protein folding chips and SNP chips using LFH technology, protein-based fluorescent micro array data, air sensing data, renewable energy data and geohazard prediction data. The development programme within the project will focus on the design and implementation of the grid infrastructure, data engineering, information structuring and knowledge discovery. Each of these components will provide mechanisms to deal with high throughput informatics. As well as delivering a practical distributed discovery plafform, the project will include the establishment of a set of standards for representing and communicating high throughput information for integrated research, to be promoted through dissemination and collaboration with related global distributed data research programmes.

Mygrid: Directly Supporting the E-Scientist

Principal Investigator: Professor Carole Goble, Dept of Computer Science, University of Manchester. Email: cgoble@cs.man.ac.uk

Collaborating Universities: Nottingham (School of Computer Science and IT), Sheffield (Dept of Computer Science), European Molecular Biology Laboratory.

Collaborating Companies: IBM UK Ltd; Epistemics Ltd; GlaxoSmithKline; SUN Microsystems Ltd; AstraZeneca UK Ltd; GeneticXchange; Network Inference; Merck KgaA

The Mygrid project will design, develop and demonstrate higher-level functionalities over an existing Grid infrastructure that supports scientists making use of complex distributed resources. The project will develop a "workbench" that supports: (i) the scientific process of experimental investigation, evidence accumulation and result assimilation; (ii) the scientist's use of the community's information; and, (iii) scientific collaboration, allowing dynamic groupings to tackle emergent research problems. The workbench will provide to the scientist personalization facilities relating to resource selection, data management and process enactment. The design and development activity will be informed by and evaluated using problems in bioinformatics, which is characterized by a highly distributed community, with many shared tool resources. Mygrid will develop two application environments, one that supports individual scientists in the analysis of functional genomic data, and another that supports the annotation of a pattern database. Both of these tasks require explicit representation of scientific processes, and have challenging performance requirements.