Date of Award

Fall 9-2014

Embargo Period

12-9-2015

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Civil and Environmental Engineering

Advisor(s)

Semiha Ergan

Abstract

Currently, corrective maintenance constitutes more than 55% of all maintenance activities in average facilities management groups in the US and. Among corrective maintenance activities, heating, ventilating and air conditioning (HVAC) related problems require prompt attention because people spend about 90% of their time inside buildings and such problems directly affect occupants’ productivity and health. However, troubleshooting of HVAC related problems is still recognized as a complex and challenging task due to the lack of apparent causes of the problems and information access/verification issues during the troubleshooting process. Three specific challenges this research addresses include: (a) the lack of knowledge of work order characteristics that result in change of possible causes of reported problems and the lack of a formal definition of domain information for troubleshooting HVAC related problems; (b) the lack of a formalism that systematically reduces the search space of possible causes, and data access issues due to dispersed data storage; and (c) the lack of efficient ways for HVAC mechanics to synthesize and comprehend information for decision-making during troubleshooting of HVAC related problems. In order to address the challenges stated above, this research aims to develop a framework which can enable identification, retrieval and visualization of the information that is required for troubleshooting of HVAC related problems. To develop such a framework, the research work I have done includes (1) identifying generic information requirements that HVAC mechanics need during troubleshooting of HVAC related problems and characteristics of work orders that affect applicable information requirements for a given work order; (2) development of representation schema and reasoning mechanisms to automatically identify applicable causes and retrieve relevant information for a given work order; and (3) evaluation of the efficiency improvement of the decision-making process during tasks of troubleshooting HVAC related problems using visualization platform. The research findings have been validated in terms of: (1) the generality of the identified information requirements of HVAC mechanics and the characteristics of work orders for troubleshooting of HVAC related problems; (2) the efficiency and accuracy of the developed formalism in identifying applicable causes and retrieve information for different work order contexts; and (3) the efficiency improvement using an integrated visualization platform in the process of troubleshooting of HVAC related problems.

Share

COinS