Roadmap

Click to proceeed to Collaborative Authoring

The Roadmap Draft has been updated and now the first four chapters can be viewed. You are invited to send your comments and suggestions to the Service Centre.
Please click HERE to proceed.

NiSIS will stimulate the emergence of new research and application themes/dimensions inspired by current rapid progress in the understanding of natural living systems. Impressive innovative techniques in microbiology during recent decades have given rise to huge data sets at genetic levels. These contain information of underlying architectures with probabilistic features, and the data have vast uncertainties caused by big noise levels. This challenge is beginning to spawn new methods of exploration and exploitation in the area of data-mining. Incredible advances in measurement and monitoring techniques sparked off by physics and chemistry have led to exciting and unprecedented understanding, but this knowledge needs a new spin-off into large information systems design and implementation. At a macroscopic level, the behaviour of animal colonies (e.g. bees and ants) provide insight into structures/societies which possess fault and damage tolerance. This is providing concepts in systems security, both for docile and hostile environments. Also, such paradigms are becoming important for large scale optimisation using principles of swarm intelligence, such as bird migratory flocking. These algorithms are better than Genetic Algorithms (GA) since they are much faster for multi-objective optimisation and are simpler to understand being based on straightforward motion vectors and dynamics.

To achieve the above mentioned objectives it is necessary to recognise and incorporate the vast but fragmented knowledge about nature-inspired processes, information systems and behaviour of living systems as well as systems in the process industry. To integrate and manage this knowledge towards the engineering of IST systems it will be necessary to move the well-established engineering design principles developed over many decades into areas which involve structure, function and behaviour which are significantly beyond the current mind-set of designers of modern high technology artefacts. It must learn from metaphors within the disciplines of the life and physical sciences, mathematical theory and social and economic systems of our modern world.

In terms of structure, the migration from conventional to nature-inspired intelligent systems will involve different morphologies to those of conventional uni-directional connectivity and modularity which give uniquely describable performance. Standard ideas of aggregation and decomposition must be replaced with flexible decentralised paradigms. Reconfigurable structures are needed to attend to the needs of self-repair/healing.

At a functional level, the well-understood engineering cycle must adjust from its linear sequencing of specification to self-goal-setting under constraints, from planning to self-forecasting, from design to ”rule-driven” self-organisation, from analysis to self-understanding, from implementation to self-realisation, from observation to self-monitoring and from control to self-decision-making/autonomy.

In the area of behaviour, we must move from fixed parameter models to those which embrace adaptive algorithmic descriptions. Behaviour itself will change from monotypical predictable modes to those which are multitypical, competitive, co-operative, collaborative, deceptive and goal-driven forms of behaviour. This will require self-aware IT networks, capable of automatic reconfiguration of topology, and self-organising information repositories forming local clusters of unstructured data. The ultimate aim is for self-designing and self-maintaining software IT systems.

All of the above require an integrated approach to system design and implementation. Engineering has vast experience in modelling of the classical sort based on dynamic equations. This knowledge needs integration into hybrid intelligent structures based on naturally-occurring structures and dynamics. They need synergetic coupling, both via empirical design and theoretical analysis.