Pattern-Based Generation of 4D-Schedules for Robust Construction Management

The creation of construction schedules is a time-consuming manual process. While project planning depends on many tasks and parameters, the quality of schedules highly relates on the engineers' experience and the access to reliable, historical project data. While knowledge applied from software can assist an engineer, the main shortcoming of current tools is the separation in generating building information and construction process models. Proposed is a knowledge-based system that automatically generates 4D-schedules based on building information models. Patterns are used to generate a process model for repetitive work flows, such as placing formwork, reinforcing, and concreting. Pattern-enriched attributes, i.e. the topology of the building structure, are automatically applied to objects in the model. Resource allocation, such as varying the performance factors or the amount of work crews, becomes possible. Such analysis can still be evaluated and exported in the format of traditional Gantt-diagrams or color-coded 4D-schedules and the construction process model can be used for detailed scheduling, construction time control, and time-based quantity evaluations. The developed approach was validated for the concreting work of a 12-story tall office building under construction. Preliminary results indicate gains in accuracy and efficiency of the developed approach compared to the traditional scheduling process.