ADVISE - Analysis of Data in a VISual Environment

By merging statistical and visualization methods throughout the exploration process, ADVISE will provide insight into the increasingly large and complex datasets that now occur routinely in many application areas. ADVISE will be service-oriented, making use of recent developments in web service technology and distributed visualization. Tailored applications will be created using the ADVISE toolkit, providing commercially viable solutions in targeted application areas such as pharmaceutics, environmental science and engineering. Building on UK strengths in statistics and visualization, the project promises to make a fundamental contribution to the emerging field of visual analytics.

The collaborators in the project are NAG, VSN International and the University of Leeds. ADVISE is funded through the Department of Trade and Industry's Technology Programme, which has been designed to help businesses work collaboratively with academic partners to develop technologies that will underpin products and services of the future.

Self-adapting software for Grid-based numerical simulation

In the past, when scientific teams have developed their simulation codes they have often done so with a particular target computer in mind (the one to which they have access) and so have selected their parallel solution algorithm, and tuned the various algorithmic parameters, specifically for this machine. With the advent of Computational Grids this approach is no longer viable since a given code will run on a variety of different architectures or, ideally, could be split between a number of different parallel computers (with different architectures). This is where the project proposed here fits in. The goal of this work is to move towards the situation whereby numerical simulation codes can be written so as to self-adapt to the architecture on which they find themselves running. This means selecting the most appropriate parallelisation strategy and automatically tuning the parameters associated with it so as to get the best possible computational performance. Furthermore if the behaviour of the computational resources or the network change during the simulation, or additional resources become available, then the software should be able to adapt dynamically to respond to these changes in its environment.  (EPSRC Reference:EP/C010027/1)

Data visualization is one of the most important areas of information management and no more so than for those "Grand Challenge" problems that require High Performance Visualization (HPV). HPV is characterised by high-quality graphics, large datasets, computationally-intensive tasks, large scale data distribution and often extensive data communication. A typical HPV task is a complex feedback process, involving data collection, visualization design, task parallelisation, immersive visual display and interfacing with the corresponding data generator such as a simulation engine. Recent hardware and software advances have demonstrated that it is now practicable to run large visual computing tasks over heterogeneous hardware with output on multiple types of display devices.

An Advanced Environment for Enabling Visual Supercomputing (EViz)

e-Viz User Interface, running an application to visualize pollution from a smoke stack

As the complexity of the enabling infrastructure increases, then so too do the demands upon the programmers for task integration, as well as the demands upon the users of the system. This places importance on system developers to create systems that reduce these demands. Such a goal is an important factor of autonomic computing, aspects of which we have used to influence our work.

In this collaborative project, we researched into the infrastructural technology for visualization from several angles. We built on our knowledge of the historic developments and applications to formulate a novel conceptual framework. We developed a simulation system to experiment with a variety of problems in an abstract manner. We designed and implemented new algorithms, techniques, modelling schemas, and systems to establish the technical feasibility of our proposed conceptual framework. We evaluated the developed systems with a collection of visualization applications.

Grid -Based Visualization and Computational Steering (gViz)

The gViz project - in collaboration with the Universities of Oxford and Oxford Brookes, CCLRC/RAL, and three industrial companies: NAG, IBM UK and Streamline - was carried out under the UK e-science core programme in the period 2002-04.  It studied visualization middleware for e-science - focusing on how to evolve visualization systems to Grid computing. Specifically it extended IRIS Explorer to work in a secure, distributed fashion, with the modules in the visualization pipeline executing on remote Grid resources.  It is expected that this software will be available in the next release of IRIS Explorer from NAG Ltd. 

Many 2D Heart simulation running on the grid, being visualized and steered from the desktop.

The project also studied computational steering in Grid environments, developing a library, the gViz library, that allows a simulation running remotely on the Grid to be monitored and controlled from the desktop.  An important contribution of gViz, due largely to Oxford Brookes, was the development of skML, an XML language to describe the syntax of dataflow visualization.  The gViz project has built on our earlier work on collaborative visualization - specifically, the EPSRC COVISA project which showed how dataflow visualization systems can be extended to multi-user working, by allowing individual dataflow pipelines to interconnect in order to share data and parameters.  This is now an integral feature within IRIS Explorer.