A knowledge engineering approach to the analysis and evaluation of schedules for mid-rise construction

A deeper understanding is needed of the process used by owners and contractors to evaluate the quality and reasonableness of construction schedules. The work described in this dissertation contributes an additional step to this understanding by articulating a subset of scheduling principles, thus enabling models of the construction schedule evaluation process to be created for subsequent automation. This research recommends that construction schedule criticism be assigned a higher project management task priority. Acknowledgment of importance to this task should further motivate large, medium, and small firms to perform schedule criticism with the same formality as is currently done by only a few large corporations. This thesis also gives guidelines as to how schedule criticism should be carried out. The methodology utilized in this research is aimed first and foremost at extracting, formalizing, and articulating a subset of judgmental knowledge about construction schedule analysis. This objective is accomplished by investigating knowledge acquisition techniques capable of eliciting and structuring an amorphous mass of scheduling knowledge. The lessons learned during this process, as well as the synthesized expertise represent major contributions of this dissertation. This study also adapts an overall knowledge metamorphosis process, the results of which represent scheduling knowledge in three different styles: (1) English form; (2) English-like notation; and (3) computer language syntax. The study concludes that the first two knowledge representation formats are particularly necessary if an operational system is to be maintained by its users. This research also makes contributions of an applications nature. The prototype systems developed endorse the proposition that knowledge-based systems technology offers programming techniques which facilitate the representation and manipulation of scheduling knowledge. The infrastructure of knowledge representation schemes provided by these prototype systems will be of value during the future development of an operational system. Such a critic system can be designed at a sufficiently generic level because there is a large body of construction scheduling knowledge that has been categorized and structured. This study does not represent a final solution by itself, so much as another step in the process of learning how computers can better assist construction managers, freeing them to tackle difficult problems.