?>

Go to content

Home
ETSIINF en Twitter ETSIINF en Facebook
Inicio > About us > Press office > Natural Computing Group

The FIUPM’s Natural Computing Group at the forefront of research into biologically-inspired computational models

The group is working in the field of membrane computing on the development of massively parallel algorithms applying evolution rules

15 January 2008. Today’s computer scientists have a new and exciting challenge: characterize and develop new models inspired primarily by nature. In this context, biology offers wide-ranging opportunities for developing what are known as non-conventional computational models.

Information technologies development has evidently contributed to the advances in biotechnology research through the massive use of algorithms and systems provided by bioinformatics and computational biology. Right back in the very early days of computing, John von Neumann proposed the development of models as cellular automata aiming to replicate key aspects of living beings, like self-reproduction.

More recently, two clearly biologically-inspired paradigms are being applied very effectively to solve applied problems in business and industry: artificial neural networks and genetic algorithms. However, there is still room for advancement. Instead of developing computing systems inspired by biological processes, biological substrates and biological processes can be used directly to codify, store and handle information, as Leonard Adelman demonstrated with his pioneering experiment on DNA computing.

Beyond cell computing

In this context, cell computing goes one step further. It is an emerging field bridging computer sciences and biology and offers a model of non-conventional computing, which, together with quantum computation, addresses complex adaptive systems. This novel field of research is partly an evolution of DNA computing, driven by results from biological systems and nanotechnology.

The recent advances in cell biology, computational biology, bioinformatics and biological systems are promising. They are helpful for furthering the understanding of the complexity of biological systems, particularly aspects related to how information is coded and how the processes that they enact to survive in dynamic and very often hostile environments are coordinated.

These processes have been specified as models that help researchers to transform cells into computational devices with sensors, internal states, transition functions, etc. Additionally, all this knowledge will allow cells to be used as nanodevices that can be programmed to develop definite tasks like generating specific drugs, managing chemicals production, etc.

At the same time, the understanding of these mechanisms is inspiring new algorithmic techniques for adaptable hardware and software architectures that can solve a broad spectrum of hard and poorly defined problems in many application areas.

The key strengths of cellular systems are energy efficiency, massive parallelism, self-recoverability, self-maintainability and evolvability. All these features make the cell a primary focus of attention in the 21st century, the century of nanoscience and information technologies.

Cell computing, then, is laying the foundations for mesoscale machines that can interact with and bridge the nanoscale and the macroscale and, at the same time, inspire new robust, distributed and evolutionary computational models.

Research into biologically-inspired computational models

The Universidad Politécnica de Madrid’s Natural Computing Group (GCN – FIUPM) has the mission to promote basic research into biologically-inspired computational models within computer sciences and artificial intelligence.

Through its membership of the European Molecular Computing Consortium, the group participates in acclaimed national or international forums or events that promote natural computing. As a member of this consortium, GCN-UPM participates in the consortium’s half-yearly meetings and all the events it sponsors in both Spain and Europe, including meetings, workshops, conferences, etc.

The GCN-FIUPM supports investigation in different fields of research related to natural computing: artificial neural networks, genetic algorithms, molecular computing (DNA computing and membrane computing, evolutionary processor networks), chaos and complex systems, etc.

In the field of membrane computing, the GCN is working actively on the development of massively parallel algorithms applying evolution rules, the development of hardware and software to deploy membrane systems and the development of dedicated hardware for these systems.

It is also researching the development of general-purpose membrane processors that can execute different alternative membrane systems whilst preserving their key characteristics: massive parallelism and non-determinism.

Three major lines of research

In this context, the group has three major lines of ongoing research:

• Design of non-deterministic and massively parallel algorithms to apply evolution rules in transitional P-systems and data structure compression for optimizing memory storage
• Design of hardware/software architectures to determine the resources required and optimize evolution times irrespective of the technologies used to deploy the membrane systems
• Deployment of the computational model with membranes at three levels: