Report : analysis of maintenance database inputs

In the previous phase of this project, 2002-059-B Case-Based Reasoning in Construction and Infrastructure Projects, demonstration software was developed using a case-base reasoning engine to access a number of sources of information on lifetime of metallic building components. One source of information was data from the Queensland Department of Public Housing relating to maintenance operations over a number of years. Maintenance information is seen as being a particularly useful source of data about service life of building components as it relates to actual performance of materials in the working environment. If a building is constructed in 1984 and the maintenance records indicate that the guttering was replaced in 2006, then the service life of the gutters was 22 years in that environment. This phase of the project aims to look more deeply at the Department of Housing data, as an example of maintenance records, and formulate methods for using this data to inform the knowledge of service lifetimes.

Machine learning techniques to improve software quality

A significant proportion of the cost of software development is due to software testing and maintenance. This is in part the result of the inevitable imperfections due to human error, lack of quality during the design and coding of software, and the increasing need to reduce faults to improve customer satisfaction in a competitive marketplace. Given the cost and importance of removing errors improvements in fault detection and removal can be of significant benefit. The earlier in the development process faults can be found, the less it costs to correct them and the less likely other faults are to develop. This research aims to make the testing process more efficient and effective by identifying those software modules most likely to contain faults, allowing testing efforts to be carefully targeted. This is done with the use of machine learning algorithms which use examples of fault prone and not fault prone modules to develop predictive models of quality. In order to learn the numerical mapping between module and classification, a module is represented in terms of software metrics. A difficulty in this sort of problem is sourcing software engineering data of adequate quality. In this work, data is obtained from two sources, the NASA Metrics Data Program, and the open source Eclipse project. Feature selection before learning is applied, and in this area a number of different feature selection methods are applied to find which work best. Two machine learning algorithms are applied to the data - Naive Bayes and the Support Vector Machine - and predictive results are compared to those of previous efforts and found to be superior on selected data sets and comparable on others. In addition, a new classification method is proposed, Rank Sum, in which a ranking abstraction is laid over bin densities for each class, and a classification is determined based on the sum of ranks over features. A novel extension of this method is also described based on an observed polarising of points by class when rank sum is applied to training data to convert it into 2D rank sum space. SVM is applied to this transformed data to produce models the parameters of which can be set according to trade-off curves to obtain a particular performance trade-off.

Teaching tribology and maintenance-related subjects : a hands-on focus

In this paper different aspects of teaching tribology and maintenance-related subjects with a hands-on focus at Queensland University of Technology (QUT) are presented and discussed. As part of the study, a combination of data from core units, such as engineering design units, and elective units, was used, in addition to laboratory experiments, real-life projects, interactive software packages and industry visits. The mechanical engineering curriculum structure used at QUT, consisting of the main specialization (first major) and the second specialization (second major), is also discussed with specific emphasis on the teaching of tribology and maintenance-related subjects. To evaluate students' satisfaction with the novel teaching approaches used, tailored questionnaires were used as well as QUT's online learning experience survey (LEX). Statistical results of these sureveys are presented and discussed. In summary, these showed that students overwhelmingly support the hands-on and practical focus in teaching tribology and maintenance-related subjects and that the teaching approaches used shorten the learning curve and make students better prepared for integration in the workplace.

Improving Usability of Software Refactoring Tools

Post-deployment maintenance and evolution can account for up to 75% of the cost of developing a software system. Software refactoring can reduce the costs associated with evolution by improving system quality. Although refactoring can yield benefits, the process includes potentially complex, error-prone, tedious and time-consuming tasks. It is these tasks that automated refactoring tools seek to address. However, although the refactoring process is well-defined, current refactoring tools do not support the full process. To develop better automated refactoring support, we have completed a usability study of software refactoring tools. In the study, we analysed the task of software refactoring using the ISO 9241-11 usability standard and Fitts' List of task allocation. Expanding on this analysis, we reviewed 11 collections of usability guidelines and combined these into a single list of 38 guidelines. From this list, we developed 81 usability requirements for refactoring tools. Using these requirements, the software refactoring tools Eclipse 3.2, Condenser 1.05, RefactorIT 2.5.1, and Eclipse 3.2 with the Simian UI 2.2.12 plugin were studied. Based on the analysis, we have selected a subset of the requirements that can be incorporated into a prototype refactoring tool intended to address the full refactoring process.

Safety management in repair, maintenance, minor alteration, and addition works : knowledge management perspective

Safety of repair, maintenance, alteration, and addition (RMAA) works have long been neglected because RMAAworks are often minute and only last for a short period of time. With rising importance of the RMAA sector in many developed societies, safety of RMAA works has begun to draw attention. Many RMAA contracting companies are small- and medium-sized enterprises (SMEs) that do not have comprehensive safety management systems. Existing safety legislation and regulations for new construction sites are not fully applicable to RMAAworks. Instead of relying on explicit and well-established safety systems, tacit safety knowledge plays an extremely important role in RMAA projects. To improve safety of RMAAworks, safety knowledge should be better managed. However, safety knowledge is difficult to capture in RMAA works. This study aims to examine safety management practices of RMAA contracting companies to see how safety knowledge of RMAA projects is managed. Findings show that RMAA contracting companies undertaking large-scale RMAA projects have more initiatives of safety management. Safety management of small-scale RMAA works relies heavily on the motivation of site supervisors and self-regulation of workers. Better tacit knowledge management improves safety performance. To enhance safety capability of RMAA contracting companies, a knowledge sharing culture should be cultivated. The government should provide assistance to SMEs to implement proper safety management practices in small-sized projects. Potentials of applying computer software technology in RMAA projects to capture, store, and retrieve safety information should be explored. Employees should be motivated to share safety knowledge by giving proper recognition to those who are willing to share.

Determination of preventive maintenance lead time using hybrid analysis

The time for conducting Preventive Maintenance (PM) on an asset is often determined using a predefined alarm limit based on trends of a hazard function. In this paper, the authors propose using both hazard and reliability functions to improve the accuracy of the prediction particularly when the failure characteristic of the asset whole life is modelled using different failure distributions for the different stages of the life of the asset. The proposed method is validated using simulations and case studies.