An automated corrective action selective assistant

• Main Author: Fayek, Aminah
• Language: English
• Published: 2008
• Online Access: http://hdl.handle.net/2429/2909

Automated daily site reporting as an integral part of a project management system has the potential to be the missing link for effective construction project monitoring and control. Problem trends over time can be detected more readily through the maintenance of daily site records that track problems occurring against activities; timely corrective action can thus be initiated more quickly. The goal of this thesis is to develop a schema to perform the automated interpretation of daily site records to identify activities that are experiencing difficulties; to identify the source(s) of these problems; to identify the types of problems resulting; to find collaborating information from the daily site records to validate the causes of these problems; and to suggest likely corrective actions. A framework is presented wherein each component of the analysis schema is defined. Fuzzy logic is used to define the imprecise relationships that exist between these components. A prototype system is developed to implement and test the schema. The components of the prototype are: a user interface; data files; an application program; an inference engine; and an expert rule base. The prototype is validated by way of several simple examples solved both manually and by the computer. A case study of a hypothetical construction project is also treated by automated interpretation to demonstrate the workability of the prototype on a multi-activity project. The model developed in this thesis would benefit from addition to and refinement of the expert rule base, and further investigation of the use of fuzzy logic to incorporate in the analysis other items of information that can be collected on site. A future system would make use of user feedback, reporting on which corrective actions have been implemented and their impacts on performance.