 Introduction
Conferences / Workshops
|
ADVANCED COMPUTATIONAL TECHNOLOGIES ™ BENEFIT FROM THE KNOWLEDGE AND EXPERIENCE OF ONE OF THE UK'S LONGEST STANDING RESEARCHERS AND PRACTITIONERS IN THE FIELD OF EVOLUTIONARY SEARCH, EXPLORATION AND OPTIMISATION. FOR THOSE UNFAMILIAR WITH THE FIELD ACT CAN PROVIDE:
A RANGE OF CONSULTANCY AND SOFTWARE DEVELOPMENT SERVICES IS AVAILABLE TO SUPPORT THE REALISATION OF THOSE ASPECTS OF THE TECHNOLOGY'S POTENTIAL THAT ARE SPECIFIC TO YOUR REQUIREMENTS. SUCH A REQUIREMENT MAY RELATE TO, FOR INSTANCE:
WHATEVER THE CHARACTERISTICS OF YOUR PARTICULAR PROBLEM AREA ACT'S BREADTH OF APPLICATION EXPERIENCE WILL ENSURE THAT AN APPROPRIATE SEARCH, EXPLORATION AND OPTIMISATION FRAMEWORK CAN BE DEVELOPED TO PROVIDE BEST UTILITY AND REALISE FULL POTENTIAL. |
Advanced Computational Technologies offers a range of services relating to the utilisation of Evolutionary and Adaptive Computing techniques. Such techniques offer powerful search and exploration capabilities for the identification of high-performance, novel and competitive solutions to industrial and commercial problems defined by many interacting variable parameters. Appropriate utilisation can ensure that search across a far wider space of possible problem solutions is achieveable within time and budget constraints.
The flexibility of the various available techniques ensures their easy integration with a wide range of problem areas and many types of problem simulation. In our experience even relatively simple application can often result in a reduction of man-hour requirement with significant savings in lead time in addition to overall product improvement. In more complex problem areas the appropriate utilisation of these techniques can provide high-performance solutions that satisfy many multiple, conflicting objectives whilst their exploratory nature can also overcome the presence of heavy, complex problem constraints ensuring the identification of optimal feasible solutions..
ACT can offer the application of these technologies to well-defined problem areas or their integration with current practice to provide support within decision-making processes. Typical potential areas of application include:
- Engineering design
- Resource scheduling and timetableing
- Facility layout
- Product design
- Data mining
- Systems identification / Response curve modelling
- Optimal loading and packing
- Structural design
- Stock production, storage and movement
- Minimisation of material usage
- Routing problems relating to pipe networks, power supply etc
- Portfolio design
- Organisational optimisation
The significant benefits offered by the approriate utilisation of these technologies is very apparent from their successful application in a wide range of problem areas. Such application has significantly improved performance whilst reducing costs relating to material, lead time and staff resource. Utilisation of the technology can also support a better understanding of the complexities of your problem areas.
ACT has extensive experience of the application and integration of a range of Evolutionary and Adaptive Computing and complementary computational intelligence techniques. These include:
- Genetic Algorithms
- Simulated Annealing
- Tabu Search
- Ant Colony Models
- Genetic Programming
- Software Agent Design
- Neural Networks
Such experience supports the rapid identification of appropriate strategies to solve your specific problems or the bespoke development of search, exploration and optimisation processes that will best integrate with current practice.
Home
Ian Parmee PhD, CEng, MInstE
Director
Advanced Computational Technologies
Harbour View
29, Embankment Rd
Kingsbridge
Devon
TQ7 4JY
Email: iparmee@ad-comtech.co.uk
Ian Parmee has some fifteen years experience of the utilisation of evolutionary and adaptive computing relating to the solution of complex industrial and commercial problems. His nine year involvement with the establishment and development of the highly successful Plymouth Engineering Design Centre (PEDC) at the University of Plymouth has seen extensive collaboration and / or consultancy with a wide range of industrial organisations. These include Rolls Royce, Rolls Royce Associates, British Energy, Lafarge Braas, British Aerospace, Perkins Technology, Unilever Research and Alsthom. He has been particularly involved in the development of evolutionary and adaptive computing strategies for generic application across a broad spectrum of application areas. Having held Directorship of the Centre for several years he has now expanded the activities of ACT to better promote and support the take-up of these technologies within the industrial and commercial sectors.
Ian Parmee now holds a Professorship at the University of the West of England, Bristol; a University which maintains a long standing and successful involvement with Evolutionary Computing and Computational Intelligence application across a diverse spectrum of industrial and commercial problem areas. Such involvement enhances, extends and supports the services that ACT can provide.
Ian Parmee is involved in a number of Networks funded by the European Community and UK Research Councils. Formerly chair of the Industrial Liaison Committee of the European Network of Excellence in Evolutionary Computation and remaining a member of the Management Committee, he now also chairs the EPSRC Network in Adaptive Computing in Design and Manufacture which comprises both industrial and academic members. He has published extensively in high-quality international journals and conference proceedings whilst contributing to and editing a number of books relating to the development and utilisation of the technologies (see publications list).
He has organised and chaired the biennial Conference 'Adaptive Computing in Design and Manufacture' since 1994. This event has always attracted international contributors and delegates most of whom are leaders in the field of evolutionary computing application. In addition, he has been actively involved in the organisation of the leading evolutionary computing conferences holding various chairs on their organising committees. The same level of involvement is apparent in a number of highly regarded engineering design and optimisation events. He sits on program committees relating to many of these conferences supporting the selection of high-quality papers and performs the same activities for a number of international journals. He has presented several invited Keynote talks at high-profile events and contributed many invited papers, tutorials and workshops to the leading conferences (see Involvement in the Field). His book 'Evolutionary and Adaptive Computing in Engineering Design' has recently been published by Springer-Verlag.
To summarise, Ian Parmee's reputation has been built upon the development of evolutionary and adaptive computing strategies and their integration with a wide range of complex real-world problem areas. He has always adopted a problem-oriented approach that has resulted in significant benefit to those organisations with whom he has been involved. Both his applied and fundamental research relating to such activity has been broadly published across the academic research communities. The dissemination of information to both industry and academia has been well supported through his involvement and active support of Networks, Workshops, Conferences and Journals relevant to the area.
Home
ACT can offer a number of services to industrial and commercial organisations. These include:
- Introductory seminars (one-to-few or one-to-many) illustrated by industrial application to raise in-house awareness of the technologies,
- Consultation relating to possible areas within your organisation that could benefit significantly from the introduction of the technology
- Application of appropriate search, exploration and optimisation algorithms to your specific multi-variate problems.
- Provision of evolutionary / adaptive software tailored to your requirement
- Integration of search, exploration and optimisation frameworks with decision-making environments.
- General systems modelling and the integration of such models with evolutionary search techniques.
Home
The technologies can offer significant adavantages over more traditional optimisation techniques which can appear inflexible when compared to the diverse search and exploration capabilities of evolutionary and adaptive computing. The common attributes of the various evolutionary and adaptive search techniques of particular relevance to practical, complex problem-solving include:
- requirement for little, if any, a priori knowledge relating to the problem. . Evolutionary and adaptive search techniques can therefore utilise a wide range of model / simulation type and structure
- excellent exploratory capabilities. The techniques initially randomly generate trial solutions and the extent of subsequent search around such solutions depends upon their relative performance. Further sampling of diverse solution sets can continue throughout the search process through the action of various operators.
- ability to avoid local optima. The stochastic nature of the various algorithms combined with continuing random sampling of the search space can prevent convergence upon local sub-optima.
- ability to handle high dimensionality. Successful application to problems described by greater than four hundred variable parameters is possible.
- robustness across a wide range of problem class. The techniques can generally outperform more deterministic optimisation algorithms across a wider range of problem classes.
- the provision of multiple good solutions. If required evolutionary and adaptive search strategies can be developed that identify multiple high-performance solutions.
- the introduction of multi-objective approaches which can be easily and successfully integrated to provide a range of high-performance compromise solutions for further off-line evaluation.
Such techniques include evolution strategies; genetic algorithms, evolutionary programming, genetic programming, ant colony models, tabu search, scatter search and simulated annealing. These techniques have become firmly established and are now becoming widely utilised within industry.
Home