ESPRIT Project: 29880

A Model For Deployment of High Performance Solutions on Internet/Intranets (ADELFI)

A project supported by:: Commission of the European Communities
DG XIII - Industry
BU 31 1/33
Rue de a Loi 200
B-1049 Brussels
Belgium

The ADELFI Project Web Pages
Comprehensive first set

1. What is ADELFI

ADELFI is a Collaborative Project supported by the European Commission, which was launched on December 1, 1998, and is due to be completed on November 30, 2000.

2. The objectives of ADELFI

2.1. General

The main objectives of the ADELFI project are:-

to develop generic software tools for the deployment of web-based:

computationally-intensive Computational Fluid Dynamics (CFD) applications, and
data-intensive Data-Mining (DM) applications;

to demonstrate the use of these tools by end-users for two particular CFD sectors, and three particular data-mining sectors, via both Internet and Intranets;
to make the tools commercially available, and to demonstrate their cost-effectivenesss.

2.2. In respect of CFD

2.2.1 The nature of CFD

[There now follows an extract from a lecture delivered at the EC-sponsored EUROTEX Workshop on Internet- and Web-based Computing Tuesday April 13th, 1999, Crowne Plaza Hotel, Dallas Market Center, Texas]

CFD, ie Computational Fluid Dynamics, is a technique for predicting quantitatively:

the performance of equipment, (e.g. automobiles, gasoline engines or electric motors) and
the behaviour of the:
- man-made (e.g. buildings, towns, bridges, aqueducts, harbours), and
- natural (e.g. winds, rivers, forest fires) environments,
in circumstances involving:
- the flow of fluids (e.g. steam , combustion gases, blood, oil) ,
- the transfer of heat (by conduction, convection or radiation),
- associated phase changes (e.g. melting, vaporisation, dissolution), and
- other concomitant effects (e.g. thermal stresses, chemical reactions, the generation of noise).

CFD requires, if its predictions are to be reliable:

user-friendly software which embodies the best that is known about the laws of physics and chemistry;

very-high-performance computer hardware;
high-level human expertise.

History

General-purpose CFD software first became commercially available in 1981, when CHAM's PHOENICS code was first released. Creare's FLUENT code followed two years later; and by now there must be a dozen or more packages offering similar capabilities.

From its beginnings in "high-tech" industry (e.g. nuclear power and aerospace), the use of CFD has spread widely into more-traditional sectors (e.g. metallurgical and food processing) and to some extent into lower-technology professions such as architecture and building services.

Obstacles to the wider use of CFD

Nevertheless, three obstacles stand in the way of its use by all who could benefit from it (e.g. town planners, environmental consultants, regulatory and law-enforcement agencies, surgeons, etc), namely:-

the cost of purchasing or leasing the software;
the cost of the high-powered computer hardware; and
the scarcity of adequately-trained and experienced personel.

A way around the obstacles

Use of the Internet provides, potentially, a way around the obstacles; for it allows the software and hardware to be used remotely, on a pay-by-use basis; and, in principle, appropriate human expertise can be obtained from anywhere in the world.

Once such a service has been made available, it is to be expected that the number of persons making use of CFD will increase rapidly, with great economic and social advantage, for it saves them the expenses of:

employing specialised staff;
purchasing expensive software;
purchasing and maintaining expensive hardware;

and it enables them to enjoy remote computer power, augmented by on-demand specialist advice.

CHAM has been engaged for several years in developing such a service, by way of:

the EC-supported MICA project, now completed;
the EC-supported ADELFI project, now starting; and
the commercial Simuserve operation, which will start to offer its services generally before the end of 1999.

[End of extract from EUROTEX lecture]

2.2.2 CFD application sectors addressed by ADELFI

As explained above, CFD is a technique which can be applied to a large number of industrial and environmental sectors; and, since the needs, language and culture of end-users in these sectors differ widely, customization of the user interface of the CFD software is essential.

Several sectors have been selected for attention in the ADELFI project, namely:-

Air and smoke movement in buildings
Flow of metal and gas in metallurgical processing plant
The cooling of electronics equipment, and
The flow of air around buildings and other obstacles, in a "virtual wind tunnel".

Sectors (a) and (b) are specifically represented in the ADELFI project by consortium members, namely INERIS and CSM; the others are being worked on by CHAM alone at the present time (but use of the software via the remote computing service will be offered to non-ADELFI end-users).

Sector (c) has been chosen because:

a significant proportion of industrial CFD activity is devoted to this application area,
the end-users can be easily identified and characterized, and
CHAM possesses pre-existing special-purpose software which it is fairly easy to adapt for operation via Internet.

Sector (d) has been chosen because it was already the subject of attention in the MICA project (as was sector (a)), so that here also the task is to adapt pre-existing customized software to remote-computing use.

2.3. In respect of Data-mining

Data mining has been characterised as the non-trivial data analysis process for identifying valid, useful and understandable patterns from databases. In essence, the goal of data mining is to derive models and ultimately knowledge from the data that is stored in routine processes today.

The term "data mining" originally referred to a selection of methods from machine learning and statistics that provide models of various kinds for data sets. The larger process from identifying an analytical task and extracting the appropriate data, via pre-processing and algorithmic analysis, up to evaluation of the results and deployment of the learned models, is also known as Knowledge Discovery from Databases (KDD)

In the Adelfi project, a methodology will be developed and implemented that allows users to apply data mining to their data. This will be achieved by providing means for importing data from remote databases, interactive pre-processing and mining of the data, and analysis of the results on-screen. The user will be able to drive the process over the network, via a Java applet or application. The computational power will be made available by using a three-tier client-server model on the basis of Enterprise JavaBeans.

Details about the contributions from Imperial College to the Adelfi project will be made available at:

http://ruby.doc.ic.ac.uk/adelfi

2.3.1 Data Mining sectors addressed by ADELFI:

Genetic Databases: finding and characterizing groups of a population (patients, animals, other organisms) with specific gene expression profile.
Specific problems are raised by the huge number of variables (up to 10,000 genes), with only some of them carrying relevant information, and the fact that measurements are not error-free, but carry experimental noise (the same experiment carried out twice will result in different data).

3. The partners

The proposer, and coordinating partner, of the project is:
CHAM, ie Concentration, Heat and Momentum Ltd.

3.1 CHAM

About CHAM
The web-site
Free Remote Computing Service

The other partners are, as a technology provider:

3.2 Imperial College

The web-site: Department of Computing, Imperial College of Science Technology and Medicine

3.3 INERIS

About INERIS
The web-site: Institut National de l'Environnement Industriel et des Risques (INERIS)

and, as end-users:

4. The progress of the project

The ADELFI project has been successfully completed at the end of May 2001. The final review meeting is on 19th Oct 2001. The final report summarises the activities carried out within the Project and the results obtained.

All the work specified in Workpackage 1-7 were successfully carried out and deliverables were submitted. The final products of the ADELFI project are the Data-mining HPC system which consists of the client software for data-mining and the data-mining HPC components, and the CFD (or PHOENICS) HPC system which consists of the CFD-Viewer and the PHOENICS HPC components. These products can be used both through Internet and corporate Intranet.

The evaluation exercise, by both data-mining end-user and CFD end-users, provided a good test for the ADELFI results. The outcome was encouraging, despite some deficiencies. The system proved itself to be operational and capable of supporting the level of activity to which it was subjected. Enhancements for commercial operation such as the network connectivity and the network response time will undoubtedly be necessary, but the groundwork has been done and the validity of the original concept demonstrated.