2014 TRUCE Summer School on Unconventional Computation
September 1-12, 2014 - Málaga, Spain.
September 1-12, 2014 - Málaga, Spain.
A two week programme, comprising lectures from leading researchers in unconventional computation, intensive workgroups, and leisure time, set in the beautiful surroundings of Andalusia, Spain.
Topics include bio- and nature-inspired computing, synthetic biology, reaction-diffusion computing, quantum computing, physical computation, evolution, and computer creativity.
The summer school will be structured as follows: mornings will be used for lectures and discussion sessions, and the afternoons will be used for independent workgroups, to be formed at the start of the school. Attendees will have the opportunity to present their ongoing research on the first day, and workgroups will then self-organize according to mutual interests. Evenings and the weekend will be left free for informal discussion, social events, etc.
Applications have now closed - thank you for your interest.
Venue: the school will be held at the Computer Science building of the University of Malaga, located at Campus de Teatinos, in the west-side of the city limits (Google Maps link). It is very well connected by bus to the city centre and to the Technology Park, and rather close to the airport and train/bus station.
Residence: it will be located in Teatinos, a lively neighborhood, full of restaurants and bars, and within walking distance from the Computer Science building.
Questions on the summer school may be addressed to the project coordinator, Prof. Martyn Amos (M.Amos@mmu.ac.uk)
Michael Lones (Heriot Watt, UK)
Biological inspiration: past, present and future.
The idea of taking inspiration from biology to inform computer science can be traced back to early visionaries such as Turing and von Neumann. Following the advent of genetic algorithms and multi-layer perceptrons in the 1970s and 80s, these ideas began to enter the mainstream of computer science. Since then, the field has continued to flourish, and has seen the development of new areas of bio-inspired computing such as artificial immune systems, genetic programming and swarm computing. Currently the field is undergoing a Cambrian explosion, with researchers taking inspiration from all corners of biology. However, is this really a useful approach? Will it lead to better understanding of how biological systems compute, or merely serve to fragment the field of bio-inspired computing? In this talk, I will discuss these and other contemporary issues faced by bio-inspired computing practitioners, and outline some of the possible paths for future research in this area. The talk will be illustrated with examples of my own research on bio-inspired metaheuristics, biochemical connectionism, and the application of bio-inspired methods within biology and medicine.
Angel Goñi-Moreno (CNB, Spain)
Biocircuit design through computational analysis.
The possibility to modify living systems following a rational design and engineering principles opens the door for building digital-like circuits with genetic material. Due to similarities in processing information between molecular-based and electronic-based devices, major efforts focus on building computational basics as oscillators, switches, counters or logic gates. By placing those devices inside cells we finally obtain machines ready to perform pre-defined tasks in a biological environment. Therefore, new applications in medicine or ecology can be tackled with them. Many internal features of specific systems are only discovered and understood by performing a strong mathematical analysis. In synthetic biology, mathematical modelling and computational simulations are essential in order to aid the design of biological parts and to de-convolute information obtained from experiments.
Zoran Konkoli (Chalmers, Sweden)
Fluctuation-dominated (reaction diffusion) kinetics in the context of unconventional computing: a friend or foe?
Diffusion-controlled reactions are ubiquitous in nature. They occur in systems ranging from matter-antimatter annihilation in the early universe to electron-hole recombination in materials. The main feature of such reactions is that the transport and the reaction times scales are clearly separated. The diffusion and reaction processes occur in an arbitrary but always traceable sequence. This is exactly the reason why the diffusion and the reaction terms can be clearly separated when describing these reactions theoretically. The dynamical behaviour associated with such reactions is extremely rich in terms of the phenomena that it produces (domain formation, excitable media etc). While the mechanisms behind the dynamics are well understood, it is much less clear whether these reactions are, and whether they can be, used to achieve some function. As an example for the former, in the living cell the reaction-diffusion dynamics is exploited to increase the efficiency of signalling pathways. Likewise, an example of the later scenario is the use of diffusion-controlled reactions for information processing. The “computational IQ” of such reactions comes partially from the non-linearity of the underlying dynamics, and partially from the fact that many reactants are involved. The fact that diffusion can be slow in mixing reacting particles can lead to kinetics where spatiotemporal fluctuations develop. In some situations, these fluctuations can be so pronounced that they can completely alter the kinetic behaviour of a system. Such kinetic behaviour is often referred to as “anomalous” or “fluctuation-dominated”. Typical situations when such kinetics occurs: (1) when the dimensionality of the system is low (from the cube, to the surface, to the line; the last being often the most pathological); (2) when the number of interacting particles is low. In both cases often everything we expect will happen turns out to be completely wrong. The question is whether one should try to engineer such behaviour to exploit it for information processing purposes, or whether it should be avoided; a friend or a foe? In the light of miniaturization of computing devices this question certainly deserves attention. The lectures will provide a balanced overview of these topics.
Irene D’Amico (York, UK)
Introduction to quantum computation.
The field of Quantum Computation has been expanding
exponentially over the last decade. At its core is the idea of finding a physical system with the right characteristics to build the "quantum computer", a device which can improve computer performances to levels unreachable by standard (i.e. "classical") computer. There are proposals for quantum computers based on semiconductors, superconductors, cold ions or atoms, molecules in a solvent, fullerenes and so on. Each of the proposals has advantages and disadvantages, and has been partially tested experimentally. The requirements to build a quantum computer are experimentally very challenging, so that the experiments performed in this area are at the very edge of modern techniques. The "quantum computer" is in fact based on the smallest possible quantum system (the two-level system or "quantum-bit") and on exquisitely quantum mechanical properties, such as state superposition. This tutorial aims to provide an introduction to this booming research
Mario de J. Pérez Jiménez (Sevilla, Spain)
Computing the natural way.
In recent years it has become clear that Nature holds the key to solve many of the relevant problems that mankind faces. No matter whether these challenges have a natural origin or not, bio-inspired computing models materialize an unconventional yet powerful approach, aiming to provide scientists with novel tools to study and make predictions on a huge variety of problems in different research domains. In particular, when dealing with real-life processes, we find that, indeed, Nature can teach us how to unveil their secrets by computing “the Natural way”. This course presents a computational bio-inspired framework to model real-life phenomena. Applications both at a “micro” (i.e. molecular mechanisms in cellular processes) and a “macro” (i.e. population dynamics of real ecosystems) level are studied.
Simon Harding (York, UK)
Evolution in materio.
It is argued that natural evolution is, par excellence, an algorithm that exploits the physical properties of materials. In this talk, I will discuss the possibility of using artificial evolution to directly exploit the properties of materials, possibly at a molecular level, in order to perform computations. Rather than building circuits from well known and understood components such as silicon transistors, evolution is used to configure “black box” systems such as liquid crystal or carbon nanotubes in order to build functional computational devices.The NASCENCE project is an ongoing EU FP7 project that aims to answer some fundamental questions in evolution in materio. What materials are useful? What type of signals can be used to configure them? What types of computation can they do? What advantages do they have over conventional systems?
In this talk, I will discuss the history of evolution in materio, the current state of art research and how it may develop in the future.
Simon Hickinbotham (York, UK)
Computational transitions in evolution.
Bio-inspired computing is one of the areas of focus for UCOMP. We need an understanding of what computing means in a biological setting, yet the scale of biological complexity makes this challenging. One of the ways we might approach this is to consider the origins of computation from the perspective of the origin of life in biological systems, where evolution is the driver of an increase in computational power and efficiency. Starting with the origin of life, this talk traces the computational properties of several key biological transitions, from a pre-biological soup, through the origin of life, the last universal common ancestor and on to modern single-celled organisms. UCOMP concepts such as embodiment, growth, self-assembly and programmability all have relevance here. By combining this new computational perspective with the more traditional view of biological evolution, we hope to gain a clearer view of how to implement bio-inspired computation in novel substrates.
Francisco Vico (Malaga, Spain)
Bio-inspired computing and computer creativity.
Computational Creativity has traditionally been considered weak Artificial Intelligence. Reasons are that creativity is considered an attribute of human intelligence (for AI); and that it would be hard to explore the neural mechanisms behind it (that's why 'weak', or 'narrow'), so we can only assess its quality by the final results. But, in this analysis, we typically take out of the equation an important element: Nature. Indeed, Nature is a great source of diversity and inspiration for all plastic and abstract arts (what is Michelangelo's David, but a copy a of a Nature's design?). From this rationale, we can approach creativity not only from mimicking human intelligence, but also from Nature's computational strategies. Evolutionary Computation will be revisited in this context, as the field most closely related to the mechanisms of biological diversity and morphogenesis. Concrete examples, like UMA's Iamus, will also be discussed.