One for all and all for one –
Conventional methods for the design of software are not well-suited for the development of ensembles. The current paradigm is based on the assumption that the developers are in possession of all relevant information – about the environment in which the new system will operate for instance – and can take this data into account in defining their design criteria. Software components are constructed so as to perform certain routines, irrespective of the wider significance of these tasks and how they contribute to the function of the system as a whole. The result is a system that is incapable of adapting to novel demands.
A new, EU-funded, interdisciplinary project named ASCENS intends to change all that. One of the major challenges facing developers of program ensembles is how to enable autonomous software components to link up into meaningful networks, interact in complex ways and produce a wide repertoire of behavior patterns. Recent advances in several related fields suggest that this challenge can now be effectively met. New service components must be developed, and new methods of communication and a new programming language designed and realized. At a more fundamental level, new formal methods, i.e. mathematically well-founded and proven procedures, need to be established which can provide the assurance that a system’s capacity for adaptation does not interfere with the performance of the basic tasks for which it was designed.
ASCENS has two major goals. One aim of the planned collaborations between researchers and design engineers is to achieve scientific breakthroughs for large-scale adaptive software systems. On the other hand, the project is designed to tackle very real problems. One possible application is the development of robot ensembles, whose members can act both autonomously and collectively. Such systems should be able to act together to overcome obstacles, for example. Professor Wirsing mentions one scenario in which they could be employed: ”Robot ensembles could be used in emergencies to complement the work of rescue teams by providing assistance to injured victims. The robots would have to be capable of reacting to a whole variety of unforeseen situations – without in any way endangering helpers or victims.“ The project coordinator hastens to add that these are dreams of the future. However, the ASCENS project has funds for 4 years of work – enough time to make significant progress towards the realization of the project’s vision.
More Information: www.ascens-ist.eu