Christos Emmanouilidis

A multiobjective evolutionary setting for feature selection and a commonality-based crossover operator

Feature selection is a common and key problem in many classification and regression tasks. It can be viewed as a multiobjective optimisation problem, since, in the simplest case, it involves feature subset size minimisation and performance maximisation. This paper presents a multiobjective evolutionary approach for feature selection. A novel commonality-based crossover operator is introduced and placed in the multiobjective evolutionary setting. This specialised operator helps to preserve building blocks with promising performance. Selection bias reduction is achieved by resampling. We argue that this is a generic approach, which can be used in many modelling problems. It is applied to feature selection on different neural network architectures. Results from experiments with benchmarking data sets are given.

Mobile solutions for engineering asset and maintenance management

Purpose – This paper seeks to present an overview of current trends in employing wireless solutions to engineering asset and maintenance management.Design/methodology/approach – The paper reviews literature and technological advances in wireless technologies and their applications in engineering asset and maintenance management from the viewpoint of assessing their potential impact, use and penetration in industrial practice. Rather than attempting a literature survey, the paper highlights emerging opportunities, benefits and potential hurdles in the adoption of mobile application solutions for industrial assets.Findings – Despite the fact that the use of advanced application solutions in manufacturing, production or process facilities takes place at a different scale, the emerging trend has already shown that mobile technologies have a great potential to redefine and re‐engineer the conventional setting for industrial asset and maintenance management. They have already begun to form the basis for offerin...

Flexible software for condition monitoring, incorporating novelty detection and diagnostics

Condition monitoring and machinery fault diagnosis are central to the implementation of efficient maintenance management strategies. They can be based on empirical modelling, which aims at associating measured data to machine conditions. Arguably, different monitoring tasks present different challenges to the maintenance engineer. This paper presents the development of a flexible software solution for condition monitoring, novelty identification and machinery diagnostics, which can easily be customised to a wide range of monitoring scenarios. Its main constituents are a number of independent software modules, such as the fault and symptom tree, the fuzzy classification module, the novelty detection and the neural network diagnostics sub-systems. It is implemented on two different applications, namely machine tool monitoring and gearbox monitoring.

Multiple-criteria genetic algorithms for feature selection in neuro-fuzzy modeling

This paper discusses the use of multicriteria genetic algorithms for feature selection in classification problems. This feature selection approach is shown to yield a diverse population of alternative feature subsets with various accuracy/complexity trade-off. The algorithm is applied to select features for performing classification with fuzzy models, and is evaluated on two real-world data sets. We discuss when multicriteria genetic algorithm feature selection is preferable to a sequential feature selection procedure, namely backwards elimination. Among the key features of the presented approach are its computational simplicity, effectiveness on real world problems and the potential it has to become a powerful tool aiding many empirical modeling and data mining processes.

Integrated E-maintenance and Intelligent Maintenance Systems

Development and acquisition of technological capabilities has become one of the major strategic requirements to excel commercially today in various business sectors. Front-end innovative technologies in conjunction with mass globalization and outsourcing of industrial operations has created a fruitful environment for technology-based growth and excellence. Different industries are in search for various technological solutions in their continuous efforts to improve performance in different parts of their businesses. Advanced solutions are often found implemented within range of application areas varying from corporate information management to logistics planning and coordination activities. In this setting, both industrial assets and business processes have been subjected to a technology-driven change process.

A layered e-maintenance architecture powered by smart wireless monitoring components

Industrial asset lifecycle management is facing pressing demands to rationalize asset usage in parallel with meeting production, quality, safety, environment and cost-efficiency constraints. The introduction of enabling technologies, within an e-Maintenance framework is a significant contributor to this end. Among the key relevant technological factors are web-based maintenance services, wireless sensing and identification technologies, data and services integration and interoperability, as well as mobile and contextualized computing. Within such a framework, a layered e-Maintenance architecture is introduced, leveraging upon the strengths of smart and wireless components in order to upgrade the maintenance-services from the low level of operations, to the higher levels of planning and decision making.

Development and Evaluation of Text Localization Techniques Based on Structural Texture Features and Neural Classifiers

This paper presents a text localization approach for binarized printed document images. Emphasis is given to the feature extraction and feature selection stages. In the former, several document structure elements and spatial features, likely to convey useful information, are extracted. In the latter, evolutionary multi-objective feature selection is employed to identify combinations of features with simultaneous good performance in terms of text localization sensitivity and specificity. The selected features are applied to a range of classifiers. Performance results over document image sets from known databases are presented, employing the classifiers with or without feature selection. The results suggest that the hybrid techniques, which utilize the classifiers in combination with the customized pre-processing, feature extraction and feature selection stages, exhibit promising performance on a range of document images.

Wireless Condition Monitoring and Embedded Novelty Detection

Wireless sensor networks are increasingly employed in a range of applications. Condition Monitoring in particular can benefit from the introduction of distributed wireless sensing solutions, operating with a high degree of autonomy. Wireless condition monitoring can extend the toolset available to the lifecycle management of engineering assets, offering ease of installation, flexibility, portability and accessibility. A significant hurdle for the adoption of wireless condition monitoring solutions in industry is related to the extent that such solutions can operate over long time periods, while providing adequate monitoring. At the application level of sensor nodes, recent programmable wireless modules are able to host intelligent services exhibiting some form of smart behavior that enables them to recognize events that deserve further attention. Thus, engineering assets equipped with embedded novelty detection capabilities exhibit some level of self-awareness, much needed to support enhanced and sustainable operation. Investing in sensor logic and its software implementations upgrades wireless modules to small model building agents capable of providing strategic advantages over static sensing infrastructures. The configurability and the programmable nature of recent maintenance management techniques and their decision support algorithms can now be ported to software agents hosted on tiny devices right in the operating environment of each asset. In this paper we review current enabling technologies, with respect to implementing ubiquitous maintenance solutions, and we study the design requirements for developing Novelty Detection techniques, as intelligent middleware components embedded on a single sensor module. On the basis of the identified design requirements, a conceptual architecture for the development of wireless sensor – embedded intelligence is outlined.

E-Maintenance: Trends, Challenges and Opportunities for Modern Industry

Abstract E-Maintenance has upraised growing interest in recent years. Equipment manufacturers seek to provide equipment-related services in an effort to increase their immunity to market variations. Furthermore, both equipment manufacturers and users are paying more attention to issues such as energy-efficiency and sustainability, in parallel with safety, quality and reduced costs. This emerging market landscape has given a boost to research in e-Maintenance, aimed at making more efficient use of resources and engineering assets. However, it seems that manufacturing companies and end users are somewhat uncertain on how to go ahead but also on which are the particular benefits they may gain with the adoption of e-Maintenance technologies. One reason is that various actors in this area of research offer somewhat differing views on the issue. This paper presents a generic view of e-Maintenance and points out both the benefits as well as the hurdles there still exist on the road to support the complete life cycle of a product with e-Maintenance, while providing a critical assessment of the current state of affairs. Based on this discussion, the paper seeks to assess how e-Maintenance will develop within the next five to ten years. Special emphasis is given on how the end users can benefit from e-Maintenance i.e. how radical the change really is and what are the financial implications related to its adoption.

Advances in Production Management Systems. Competitive Manufacturing for Innovative Products and Services

The two volumes IFIP AICT 397 and 398 constitute the thoroughly refereed post-conference proceedings of the International IFIP WG 5.7 Conference on Advances in Production Management Systems, APMS 2012, held in Rhodes, Greece, in September 2012. The 182 revised full papers were carefully reviewed and selected for inclusion in the two volumes. They are organized in 6 parts: sustainability; design, manufacturing and production management; human factors, learning and innovation; ICT and emerging technologies in production management; product and asset lifecycle management; and services, supply chains and operations.