The investigation on automating the performance assessment of engineering asset management for buildings and plants in their maintenance information management systems

探求工程資產管理的廠間及建築物維修紀錄系統自動化效能審查

Student thesis: Doctoral Thesis

View graph of relations

Author(s)

  • Kui Shun FUNG

Detail(s)

Awarding Institution
Supervisors/Advisors
Award date15 Jul 2014

Abstract

With the global escalating customer expectation of services and products in the Engineering Asset Management (EAM) trade, a standard for systematic EAM system was developed [BSI PAS 55, 2004]. The International Standards Organization (ISO) has adopted the PAS 55 for their development of asset management standard - ISO55000 scheduled to be launched in 2014. At present in Hong Kong, the PAS 55 has been adopted by a number of public utility corporations who provide vital community services that involve large scale investment of public services and high demand on reliability in EAM. However, in the private sector, especially in buildings and plants related operations and maintenance (O&M) services, which is a substantial portion of the EAM trade, has yet not adopted PAS 55 as their EAM systems. Another deficiency in PAS 55 is that it only provides the 'do' list of EAM but not the method or standard on how to assess or score their EAM performance in industries. This research explored the levels of PAS 55 adoption of EAM systems in the private sector and showed that EAM systems for different categories of core business in the private sector may benefit from a different level of PAS 55. To investigate the degree of adoption of the EAM practice in current industry of Hong Kong, a questionnaire, which was designed according to the criteria set out in Section 4.4.6 of PAS 55, was used to survey the performance of thirty one EAM practitioners in their information management. The original scope of coverage in PAS 55 is substantial and cannot be investigated by one thesis work. Hence, Section 4.4.6 on information management was selected as the investigation target. Another reason to select Section 4.4.6 as the main target for the survey is because information management is a very significant service business in Hong Kong. The related business includes information commonly used for banking, communications, trading and public services, etc... The questionnaire sought to survey the virtual adoption levels according to [BSI PAS 55, 2008] by mapping their EAM systems to the framework specified in the PAS 55. The design of the questionnaire is intended not only to reveal the general practice of EAM in Hong Kong, but also to assess the performance after adopting such EAM practice. Five different categories of building operations and maintenance practitioners were selected and invited to participate in the survey. The survey results were analyzed by a software tool [IBM SPSS, 2009] for checking on the validity levels of the collected data. Afterward, two artificial intelligence (AI) methods, namely the linear discrimination and the k-nearest moderation by least squares distances, were used to evaluate and model the degree of differences in O&M performance between the current practices of the EAM practitioners and those specified in the PAS 55. The AI methods are able to automate findings of the most suitable practice in EAM used by each category of buildings and plants O&M practitioners and their levels of compliance according to the PAS 55 requirements. The most suitable practice will then form a proper benchmarking level of EAM for each type of category. Achievable means have also been recommended so that buildings and plants O&M practitioners can find their most suitable ways to accomplish international recognitions in EAM and vital references in benchmarking their performance with other worldwide recognized EAM practitioners. Besides the above achievements, this thesis also reports the research work on the creation of an e-Manager system as a demonstrator so that certain categories of practitioner companies may easily set up a preset information management system to suit their needs matching with the PAS 55 framework from the internet. The e-Manager system also states the subsequent operations, which contain the know-how and the question & answer in information management commonly used in Hong Kong for any particular type of company to follow via the internet cloud-base information system. Hence, EAM practitioners may use this system to cater their clients' normal O&M. By continually using such an e-Manager system, the valuable information management experience can be accumulated. Hence, the operating database of an e-Manager system can be gradually tracked according to the PAS 55 framework. Subsequently, a particular category of EAM practitioners using this database can evaluate their information management systems mapped with the suggested PAS 55 benchmarking levels assessed by the AI methods. In future, more sophisticated AI methods, like the least squares based Support Vector Machine (SVM) and Fuzzy Logic Averaging, can be used to enhance the intelligence of the e-Manager and increase the accuracy of selecting the most suitable practice ofEAM in each category of EAM practitioners operating in Hong Kong. The current e-Manager system equipped with the AI methods only acts as a demonstrator to show the feasibility of automating scoring assessment process of EAM performance that is missing in the current PAS 55. In the future, with the help of such system, human experts who are expensive to hire and rare to find, may not be required to conduct the necessary assessment prior to the subsequent granting of certification according to PAS 55. The system is able to show how to assess a good building O&M practitioner in information management. Moreover, standardized benchmarking levels of a particular type of O&M information management will be demonstrated by this automating system. Consequently a particular building O&M practitioner can find his performance score as compared to the benchmarking levels. Most importantly, these benchmarking levels are internationally recognized and can obtain certification in EAM according to PAS 55. Since the work showed here is only related to information management in EAM, more intensive work should be conducted in the future to cover the entire scope of PAS 55, so that the gap of lacking proper assessment solution as stated in the current PAS 55 can be bridged.

    Research areas

  • Automation, Plant maintenance, Maintenance, Management information systems, Management, Evaluation, Buildings