1 A Virtual Mining Environment for Providing Dynamic Decision Support for Building Maintenance Rabee M. Reffat 1 and John S. Gero 2 1 Architecture Department, KFUPM, Dhahran 31261, Saudi Arabia http://faculty.kfupm.edu.sa/arch/rabee/ 2 Key Centre of Design Computing and Cognition, University of Sydney, NSW 2006, Australia http://www.arch.usyd.edu.au/~john Abstract. A virtual mining environment aims to provide dynamic decision support to improve building life-cycle modelling and management. This paper presents the system architecture of a virtual mining environment, its interfaces and a user scenario. This virtual mining environment integrates data mining with agent-based technology, database management systems, object-based CAD systems, and 3D virtual environments. A system prototype has been developed and implemented to support the automated feed back for building life cycle modelling, planning and decision-making. Keywords. Data Mining; Virtual Environment; Dynamic Decision Support; Building Maintenance. Introduction In building maintenance, the management of building facilities is essential to achieve better reliability and availability of the equipment installed. It is important to minimize downtime of all equipments that are used as the downtime will impact on the profitability of the building itself. Maintenance activity, as part of a business process within an internal organization, contributes to the successful operation of the physical asset. The maintenance budget is a significant cost and is thus very important for the overall economic results. It has been shown that major factors contributing to construction quality problems include inadequate information and poor communication (Arditi et al. 1998; Burati et al. 1992). The detection of previously undiscovered patterns in building maintenance systems data can be used to determine factors such as the cost effectiveness and expected failure rates of assorted building materials or equipment in varying environments and circumstances. These factors are important throughout the life cycle of a building, and such information could be used in the design, construction, refurbishment, and maintenance of a building , potentially leading to a substantial decrease in cost and increase in reliability. Such knowledge is significant for saving resources in construction projects. As the construction industry adapts to new computer technologies computerized design aids, construction, and maintenance data are all becoming increasingly available. The growth of many business, government, and scientific databases has begun to far outpace an individual’s ability to interpret and digest the data. Such volumes of data clearly overwhelm the traditional methods of data analysis such as spreadsheets and ad-hoc queries. The traditional methods can create informative reports from data, but cannot analyse the contents of those reports. A significant need exists for the application of new techniques and tools to automatically assist humans in analysing the increasing volume of data for useful knowledge. The increasing use of databases to store information about facilities, their use, and their maintenance provides the background and platform for the use of data mining techniques for future projections. The current technology for facility maintenance uses databases to keep track of information and for notification of maintenance schedules. These databases
