Germán's Publications

ABSTRACT: The new generation of ICT solutions applied to the monitoring, adaptation, simulation and optimisation of factories are key enabling technologies for a new level of manufacturing capability and adaptability in the context of Industry 4.0. Given the advances in sensor technologies, factories, as well as machine tools can now be sensorised, and the vast amount of data generated can be exploited by intelligent information processing techniques such as machine learning. This paper presents an online tool wear classification system built in terms of a monitoring infrastructure, dedicated to perform dry milling on steel while capturing force signals, and a computing architecture, assembled for the assessment of the flank wear based on deep learning. In particular, this approach demonstrates that a big data analytics method for classification applied to large volumes of continuously-acquired force signals generated at high speed during milling responds sufficiently well when used as an indicator of the different stages of tool wear. This research presents the design, development and deployment of the system components and an overall evaluation that involves machining experiments, data collection, training and validation, which, as a whole, has shown an accuracy of 78%.

ABSTRACT: This paper presents an in-process tool wear prediction system, which uses a force sensor to monitor the progression of the tool flank wear and machine learning (ML), more specifically, a Convolutional Neural Network (CNN) as a method to predict tool wear. The proposed methodology is experimentally illustrated using milling as a test process. The experiments are conducted using dry machining with a non-coated ball endmill and a stainless steel workpiece. The measurement of the flank wear is carried on in-situ utilising a digital microscope. The ML model predictions are based on an experience database which contains all the data of the precedent experiments. The proposed in-process tool wear prediction system will be reinforced later by an adaptive control (AC) system that will communicate continuously with the ML model to seek the best adjustment of feed rate and spindle speed that allows the optimization of the flank wear and extend the tool life. The AC model decisions are based on the prediction delivered by the ML model and on the information feedback provided from the force sensor, which captures the change in the cutting forces as a function of the progression of the flank wear. In this work, only the ML model component for the estimation of tool wear based on CNNs is demonstrated. The proposed methodology has shown an estimated accuracy of 90%. Additional experiments will be performed to confirm the repetitiveness of the results and also extend the measurement range to improve accuracy of the measurement system.

ABSTRACT:Advanced digitalization together with the rise of cloud technologies is a key enabler for a fundamental paradigm shift known as Industry 4.0 which pro-poses the integration of the new generation of ICT solutions for the monitoring, adaptation, simulation and optimization of factories. With the democratization of sensors, assembly systems can now be sensorized and the data generated by these devices can be exploited, for instance, to monitor their utilization, operations and maintenance. However, analyzing the vast amount of generated data is resource demanding both in terms of computing power and network bandwidth, especially when dealing with real-time changes to product, process and resource domains. This paper presents a novel cloud-based analytics framework for the management and analysis of assembly systems. It brings together standard open source tech-nologies and the exploitation of cloud computing which as a whole can be adapted to and deployed on different cloud providers, thereby reducing infra-structure costs, minimizing deployment difficulty and providing on-demand ac-cess to virtually infinite computing power, storage and network resources.

ABSTRACT: Industry 4.0 proposes the integration of the new generation of ICT solutions for the monitoring, adaptation, simulation, and optimisation of factories. With the democratization of sensors and actuators, factories and machine tools can now be sensorized and the data generated by these devices can be exploited, for instance, to optimize the utilization of the machines as well as their operation and maintenance. However, analysing the vast amount of data generated is resource demanding both in term of computing power and network bandwidth, thus requiring highly scalable solutions. This paper presents a novel big data platform for the management of machine generated data in the cloud. It brings together standard open source technologies which can be adapted to and deployed on different cloud infrastructures, hence reducing costs, minimising deployment difficulty and providing on-demand access to a virtually infinite set of computing, storage and network resources.

ABSTRACT: In the past decades many definitions of complexity have been proposed. Most of these definitions are based either on Shannon's information theory or on Kolmogorov complexity; these two are often compared but very few studies integrate both ideas. In this paper we introduce a new measurement of complexity that build on both of these theories. As a demonstration of the concept, the technique is applied to Elementary Cellular Automata and simulations of the selforganization of porphyrin molecules.

ABSTRACT: Biological and other natural processes have always been a source of inspiration for computer science and information technology. Many emerging problem solving techniques integrate advanced evolution and cooperation strategies, encompassing a range of spatio-temporal scales for visionary conceptualization of evolutionary computation. This book is a collection of research works presented in the VI International Workshop on Nature Inspired Cooperative Strategies for Optimization (NICSO) held in Canterbury, UK. Previous editions of NICSO were held in Granada, Spain (2006 & 2010), Acireale, Italy (2007), Tenerife, Spain (2008), and Cluj-Napoca, Romania (2011). NICSO 2013 and this book provides a place where state-of-the-art research, latest ideas and emerging areas of nature inspired cooperative strategies for problem solving are vigorously discussed and exchanged among the scientific community. The breadth and variety of articles in this book report on nature inspired methods and applications such as Swarm Intelligence, Hyper-heuristics, Evolutionary Algorithms, Cellular Automata, Artificial Bee Colony, Dynamic Optimization, Support Vector Machines, Multi-Agent Systems, Ant Clustering, Evolutionary Design Optimisation, Game Theory and other several Cooperation Models.

ABSTRACT: Self-assembly is a phenomenon observed in nature at all scales where autonomous entities build complex structures far from equilibrium, without external influences nor master plan. Modelling such entities and programming correct interactions among them is crucial for controlling the manufacture of desired complex structures. This work focuses on a programmability model for non DNA-based molecules and complex behaviour analysis of their self-assembled conformations. In particular, we look into modelling, programming and simulation of porphyrin molecules self-assembly and apply Kolgomorov complexity-based techniques to classify and assess simulation results in terms of information content. The analysis focuses on phase transition, clustering, variability and parameter discovery which as a whole pave the way to the notion of complex systems programmability.

ABSTRACT: Porphyrins are molecular units with fourfold symmetry and suitable for solid substrate deposition. The chemical structure of a porphyrin molecule reveals four structural units which can be synthesised with different substituent functional groups. The adequate selection of functional groups plays a central role in defining the correct intermolecular bindings that lead to a precisely tuned spatial self-assembled pattern. In this paper we explore the state-space of self-assembled programmable patterns. This is done by modelling the porphyrins molecular units using a kinetic Monte Carlo approach. Furthermore we analyse our simulations by both deriving discrete computational automata and in terms of algorithmic information content.

ABSTRACT: In a previous work we have reported on the evolutionary design optimisation of self-assembling Wang tiles capable of arranging themselves together into a target structure. Apart from the significant findings on how selfassembly is achieved, nothing has been yet said about the efficiency by which individuals were evolved. Specially in light that the mapping from genotype to phenotype and from this to fitness is clearly a complex, stochastic and non-linear relationship. One of the most common procedures would suggest running many experiments for different configurations followed by a fitness comparison, which is not only time-consuming but also inaccurate for such intricate mappings. In this paper we aim to report on a complementary dual assessment protocol to analyse whether our genetic algorithm, using morphological image analyses as fitness function, is an effective methodology. Thus, we present here fitness distance correlation to measure how effectively the fitness of an individual correlates to its genotypic distance to a known optimum, and introduce clustering as a mechanism to verify how the objective function can effectively differentiate between dissimilar phenotypes and classify similar ones for the purpose of selection.

ABSTRACT: In a previous work we have reported on the evolutionary design optimisation of self-assembling Wang tiles. Apart from the achieved findings [11], nothing has been yet said about the effectiveness by which individuals were evaluated. In particular when the mapping from genotype to phenotype and from this to fitness is an intricate relationship. In this paper we aim to report whether our genetic algorithm, using morphological image analyses as fitness function, is an effective methodology. Thus, we present here fitness distance correlation to measure how effectively the fitness of an individual correlates to its genotypic distance to a known optimum when the genotype-phenotype-fitness mapping is a complex, stochastic and non-linear relationship.

ABSTRACT: Many aspects of Nature, Biology or even from Society have become part of the techniques and algorithms used in computer science or they have been used to enhance or hybridize several techniques through the inclusion of advanced evolution, cooperation or biologically based additions. The previous NICSO workshops were held in Granada, Spain, 2006, Acireale, Italy, 2007, and in Tenerife, Spain, 2008. As in the previous editions, NICSO 2010, held in Granada, Spain, was conceived as a forum for the latest ideas and the state of the art research related to nature inspired cooperative strategies. The contributions collected in this book cover topics including nature-inspired techniques like Genetic Algorithms, Evolutionary Algorithms, Ant and Bee Colonies, Swarm Intelligence approaches, Neural Networks, several Cooperation Models, Structures and Strategies, Agents Models, Social Interactions, as well as new algorithms based on the behaviour of fireflies or bats.

ABSTRACT: The current investigations on hyper-heuristics design have sprung up in two different flavours: heuristics that choose heuristics and heuristics that generate heuristics. In the latter, the goal is to develop a problem-domain independent strategy to automatically generate a good performing heuristic for the problem at hand. This can be done, for example, by automatically selecting and combining different low-level heuristics into a problem specific and effective strategy. Hyper-heuristics raise the level of generality on automated problem solving by attempting to select and/or generate tailored heuristics for the problem at hand. Some approaches like genetic programming have been proposed for this. In this paper, we explore an elegant nature-inspired alternative based on self-assembly construction processes, in which structures emerge out of local interactions between autonomous components. This idea arises from previous works in which computational models of self-assembly were subject to evolutionary design in order to perform the automatic construction of user-defined structures. Then, the aim of this paper is to present a novel methodology for the automated design of heuristics by means of self-assembly.

ABSTRACT: Developing a problem-domain independent methodology to automatically generate high performing solving strategies for specific problems is one of the challenging trends on hyper-heuristics design. Designing hyper-heuristics is important because they raise the level of generality on automated problem solving by attempting to select and/or generate tailored heuristics for the problem at hand. In this paper, we present a three-steps methodology that combines multiple sequence alignment and grammatical induction in order to automatically generate high performing solving strategies for a combinatorial optimisation problem. We present proof-of-concept results of applying this methodology to instances of the well-known symmetric TSP. The goal here is to demonstrate feasibility rather than compete with state of the art TSP solvers. This TSP is chosen only because it is an easy to state and well known problem.

ABSTRACT: The current research trends on hyper-heuristics design have sprung up in two different flavours: heuristics that choose heuristics and heuristics that generate heuristics. In the latter, the goal is to develop a problem-domain independent strategy to automatically generate a good performing heuristic for specific problems, that is, the input to the algorithm are problems and the output are problem-tailored heuristics. This can be done, for example, by automatically selecting and combining different low-level heuristics into a problemspecific and effective strategy. Thus, hyper-heuristics raise the level of generality on automated problem solving by attempting to select and/or generate tailored heuristics for the problem in hand. Some approaches like genetic programming have been proposed for this. In this paper, we report on an alternative methodology that sheds light on simple methodologies that efficiently cooperate by means of local interactions. These entities are seen as building blocks, the combination of which is employed for the automated manufacture of good performing heuristic search strategies.We present proof-of-concept results of applying this methodology to instances of the well-known symmetric TSP. The goal here is to demonstrate feasibility rather than compete with state of the art TSP solvers. This TSP is chosen only because it is an easy to state and well known problem.

ABSTRACT: Self-assembly and self-organisation are natural construction processes where the spontaneous formation of aggregates emerges throughout the progressive interplay of local interactions among its constituents. Made upon cooperative self-reliant components, selfassembly and self-organising systems are seen as distributed, not necessarily synchronous, autopoietic mechanisms for the bottom-up fabrication of supra-structures. The systematic understanding of how nature endows these autonomous components with sufficient “intelligence” to combine themselves to form useful aggregates brings challenging questions to science, answers to which have many potential applications in matters of life and technological advances. It is for this reason that the investigation to be presented along this thesis focuses on the automated design of self-assembly and self-organizing systems by means of artificial evolution. Towards this goal, this dissertation embodies research on evolutionary algorithms applied to the parameters design of a computational model of self-organisation and the components design of a computational model of self-assembly. In addition, an analytical assessment combining correlation metrics and clustering, as well as the exploration of emergent patterns of cooperativity and the measurement of activity across evolution, is made. The results support the research hypothesis that an adaptive process such as artificial evolution is indeed a suitable strategy for the automated design of self-assembly and self-organising systems where local interactions, homogeneity and both stochastic and discrete models of execution play a crucial role in emergent complex structures.

ABSTRACT: Self-assembly and self-organisation are natural construction processes where the spontaneous Self-assembly is a distributed, asynchronous mechanism that is pervasive across natural systems where hierarchical complex structures are built from the bottom-up. The lack of a centralised master plan, no external intervention, and preprogrammed interactions among entities are within its most relevant and technologically appealing properties. This paper tackles the self-assembly Wang tiles designability problem by means of artificial evolution. This research is centred in the use of tiles that are extended with rotation and probabilistic motion, and an evolutionary algorithm using the Morphological Image Analyses method as a fitness function. The obtained results support this approach as a successful engineering mechanism for the computer-aided design of self-assembled patterns.

ABSTRACT: Cellular automata (CA) are an important modelling paradigm in the natural sciences and an extremely useful approach in the study of complex systems. Homogeneity, massive parallelism, local cellular interactions and both synchronous and asynchronous models of rule execution are some of their most prominent features, allowing scientists to model and understand a variety of phenomena in, to name but a few, the physical, chemical, biological, social and information sciences. An ubiquitous problem related with the study of complex systems by means of CA is that of parameter identification. In some cases, analytical methods are available but in many others, due to the bottom-up complexity of the underlying processes, the best route for CA identification is through design optimization by means of a metaheuristic, such as an evolutionary algorithm. In this work we report on a systematic methodology we have developed to control the spatio-temporal behavior of a CA in order to obtain a ‘designoid’ target pattern. Four independent CA-based complex systems were used to assess our hypothesis which combines clustering, fitness distance correlation and evolutionary algorithms.

ABSTRACT: Self-Assembly is a powerful autopoietic mechanism ubiquitous throughout the natural world. It may be found at the molecular scale and also at astronomical scales. Self-assembly power lays in the fact that it is a distributed, not-necessarily synchronous, control mechanism for the bottom-up manufacture of complex systems. Control of the assembly process is shared across a myriad of elemental components, none of which has either the storage or the computation capabilities to know and follow a master plan for the assembly of the intended system. In this paper we present an evolutionary algorithm which is capable of programming the so called “Wang Tiles” for the self-assembly of two-dimensional squares.

ABSTRACT: “Quorum Sensing” has been identified as one of the most consequential microbiology discoveries of the last 10 years. Using Quorum Sensing bacterial colonies synchronize gene expression and phenotype change allowing them, among other things, to protect their niche, coordinate host invasion and bio-film formation. In this contribution we briefly describe the elementary microbiology background and present a P-systems based model for Quorum Sensing which includes environmental rules and a topological representation.

ABSTRACT: “Quorum Sensing” has been identified as one of the most consequential microbiology discoveries of the last 10 years. Using Quorum Sensing bacterial colonies synchronize gene expression and phenotype change allowing them, among other things, to protect their niche, coordinate host invasion and bio-film formation. In this contribution we briefly describe the elementary microbiology background and comment on some of the approaches that have been used to model this important phenomenon. We also informally show that Quorum Sensing is computational complete and suggest some applications where this computational method could impact other areas of computer science.