Jennifer A. Harding

Data Mining in Manufacturing: A Review

The paper reviews applications of data mining in manufacturing engineering, in particular production processes, operations, fault detection, maintenance, decision support, and product quality improvement. Customer relationship management, information integration aspects, and standardization are also briefly discussed. This review is focused on demonstrating the relevancy of data mining to manufacturing industry, rather than discussing the data mining domain in general. The volume of general data mining literature makes it difficult to gain a precise view of a target area such as manufacturing engineering, which has its own particular needs and requirements for mining applications. This review reveals progressive applications in addition to existing gaps and less considered areas such as manufacturing planning and shop floor control.

An intelligent information framework relating customer requirements and product characteristics

Abstract Market driven strategies encourage enterprises to produce products that customers want to buy, and therefore can improve an enterprise’s market position. Few organisations make effective use of market, competitor and customer information. Information modelling and intelligent support tools help define product specifications focused on fulfilling customer requirements and facilitating information sharing between members of extended design teams. Design effort can be targeted at particular product features, which yield maximum benefits for customer satisfaction. The Market Driven Design System provides comprehensive, intelligent support, meeting the challenges of effectively modelling, using and sharing valuable, yet imprecise, non-technical market information during product design.

Manufacturing system engineering ontology for semantic interoperability across extended project teams

Communication, knowledge sharing and awareness of available expertise are complex issues for any multidiscipline team. Complexity increases substantially in extended enterprise environments. The concepts of an MSE Moderator have previously been considered in environments with shared information models and vocabularies. These concepts are now translated to the realm of extended enterprises, where inevitably, individual partners will have their own terminology and information sources. An MSE Ontology is proposed to enable the operation of an extended enterprise MSE Moderator to provide common understanding of manufacturing-related terms, and therefore to enhance the semantic interoperability and reuse of knowledge resources within globally extended manufacturing teams.

A model-driven ontology approach for manufacturing system interoperability and knowledge sharing

The requirements for the interoperability of semantics and knowledge have become increasingly important in Product Lifecycle Management (PLM), in the drive towards knowledge-driven decision support in the manufacturing industry. This article presents a novel concept, based on the Model Driven Architecture (MDA). The concept has been implemented under the Interoperable Manufacturing Knowledge Systems (IMKS) project in order to understand the extent to which manufacturing system interoperability can be supported using radically new methods of knowledge sharing. The concept exploits the capabilities of semantically well-defined core concepts formalised in a Common Logic-based ontology language. The core semantics can be specialised to configure multiple application-specific knowledge bases, as well as product and manufacturing information platforms. Furthermore, the utilisation of the expressive ontology language and the generic nature of core concepts help support the specification of system mechanisms to enable the verification of knowledge across multiple platforms. An experimental demonstration, using a test case based on the design and manufacture of an aerospace part, has been realised. This has led to the identification of several benefits of the approach, its current limitations as well as the areas to be considered for further work.

Textual data mining for industrial knowledge management and text classification: A business oriented approach

Textual databases are useful sources of information and knowledge and if these are well utilised then issues related to future project management and product or service quality improvement may be resolved. A large part of corporate information, approximately 80%, is available in textual data formats. Text Classification techniques are well known for managing on-line sources of digital documents. The identification of key issues discussed within textual data and their classification into two different classes could help decision makers or knowledge workers to manage their future activities better. This research is relevant for most text based documents and is demonstrated on Post Project Reviews (PPRs) which are valuable source of information and knowledge. The application of textual data mining techniques for discovering useful knowledge and classifying textual data into different classes is a relatively new area of research. The research work presented in this paper is focused on the use of hybrid applications of text mining or textual data mining techniques to classify textual data into two different classes. The research applies clustering techniques at the first stage and Apriori Association Rule Mining at the second stage. The Apriori Association Rule of Mining is applied to generate Multiple Key Term Phrasal Knowledge Sequences (MKTPKS) which are later used for classification. Additionally, studies were made to improve the classification accuracies of the classifiers i.e. C4.5, K-NN, Naive Bayes and Support Vector Machines (SVMs). The classification accuracies were measured and the results compared with those of a single term based classification model. The methodology proposed could be used to analyse any free formatted textual data and in the current research it has been demonstrated on an industrial dataset consisting of Post Project Reviews (PPRs) collected from the construction industry. The data or information available in these reviews is codified in multiple different formats but in the current research scenario only free formatted text documents are examined. Experiments showed that the performance of classifiers improved through adopting the proposed methodology.

An enterprise modeling and integration framework based on knowledge discovery and data mining

This paper deals with the conceptual design and development of an enterprise modeling and integration framework using knowledge discovery and data mining. First, the paper briefly presents the background and current state-of-the-art of knowledge discovery in databases and data mining systems and projects. Next, enterprise knowledge engineering is dealt with. The paper suggests a novel approach of utilizing existing enterprise reference architectures, integration and modeling frameworks by the introduction of new enterprise views such as mining and knowledge views. An extension and a generic exploration of the information view that already exists within some enterprise models are also proposed. The Zachman Framework for Enterprise Architecture is also outlined versus the existing architectures and the proposed enterprise framework. The main contribution of this paper is the identification and definition of a common knowledge enterprise model which represents an original combination between the previous projects on enterprise architectures and the Object Management Group (OMG) models and standards. The identified common knowledge enterprise model has therefore been designed using the OMGs Model-Driven Architecture (MDA) and Common Warehouse MetaModel (CWM), and it also follows the RM-ODP (ISO/OSI). It has been partially implemented in Java™, Enterprise JavaBeans (EJB) and Corba/IDL. Finally, the advantages and limitations of the proposed enterprise model are outlined.

Product design and manufacturing process improvement using association rules

Abstract Modern manufacturing systems equipped with computerized data logging systems collect large volumes of data in real time. The data may contain valuable information for operation and control strategies as well as providing knowledge of normal and abnormal operational patterns. Knowledge discovery in databases can be applied to these data to unearth hidden, unknown, representable, and ultimately useful knowledge. Data mining offers tools for discovery of patterns, associations, changes, anomalies, rules, and statistically significant structures and events in data. Extraction of previously unknown, meaningful information from manufacturing databases provides knowledge that may benefit many application areas within the enterprise, for example improving design or fine tuning production processes. This paper examines the application of association rules to manufacturing databases to extract useful information about a manufacturing systems capabilities and its constraints. The quality of each identified rule is tested and, from numerous rules, only those that are statistically very strong and contain substantial design information are selected. The final set of extracted rules contains very interesting information relating to the geometry of the product and also indicates where limitations exist for improvement of the manufacturing processes involved in the production of complex geometric shapes.

Personalised online sales using web usage data mining

Practically every major company with a retail operation has its own web site and online sales facilities. This paper describes a toolset that exploits web usage data mining techniques to identify customer Internet browsing patterns. These patterns are then used to underpin a personalised product recommendation system for online sales. Within the architecture, a Kohonen neural network or self-organizing map (SOM) has been trained for use both offline, to discover user group profiles, and in real-time to examine active user click stream data, make a match to a specific user group, and recommend a unique set of product browsing options appropriate to an individual user. Our work demonstrates that this approach can overcome the scalability problem that is common among these types of system. Our results also show that a personalised recommender system powered by the SOM predictive model is able to produce consistent recommendations.

The application of STEP-NC using agent-based process planning

Computer aided part programming is an information intensive activity, and provides an interesting area for the application of artificial intelligence, knowledge-based systems and recently for agent technologies. Intelligent software agents can behave and respond autonomously to set criteria. The flexibility of agents and their ability to make autonomous/cooperative decisions makes them an excellent tool in the development of future computer aided process planning systems. In addition, a new ISO standard for NC machine tool programming is emerging called ISO 14649, informally known as STEP-NC. This standard is aimed to replace the existing ISO 6983 (G&M code) standard, which has been in use since the first NC machine tools were developed in the 1950s. STEP-NC does not describe tool movements but uses a higher-level language to define operations known as Workingsteps. A new breed of intelligent controller is envisaged to determine optimized tool trajectories allowing advanced inter-operability and a bi-directional information flow which allows edits made at the machine to be fed back to the CAD/CAM model. This paper presents a STEP-NC compliant computational environment used to demonstrate agent-based process planning, resulting in the generation of STEP-NC code. The system comprises a multi-agent framework, where agents represent the individual features of the component and work independently and cooperatively to generate process plans for discrete component manufacture. This is demonstrated through a series of case study components which evaluate various feature interaction scenarios.

Towards a formal manufacturing reference ontology

Due to the advancement in the application of Information and Communication Technology (ICT), manufacturing industry and its many domains employ a wide range of different ICT tools. To be competitive, industries need to communicate effectively within and across their many system domains. This communication is hindered by the diversity in the semantics of concepts and information structures of these different domain systems. Whilst international standards provide an effective route to information sharing within narrowly specified domains, they are themselves not interoperable across the wide range of application domains needed to support manufacturing industry due to the inconsistency of concept semantics. Formal ontologies have shown promise in removing interpretation problems by computationally capturing the semantics of concepts, ensuring their consistency and thus providing a verifiable and shared understanding across multiple domains. The research work reported in this paper contributes to the development of formal reference ontology for manufacturing, which is envisaged as a key component in future interoperable manufacturing systems. A set of core manufacturing concepts are identified and their semantics have been captured in formal logic based on exploiting and extending existing standards’ definitions, where possible combined with an industrial investigation of the concepts required. A successful experimental investigation has been conducted to verify the application of the ontology based on the interaction between concepts in the design and manufacturing domains of an aerospace component.