Can algorithms support the specification of construction schedules?

Construction processes require preparation phases where rules, agreements and arrangements for the execution of the project are discussed and determined. Detailed information is worked out where among others the construction activities are named and follow up charts are determined. In practice, follow up charts are developed that describe the project in different levels of detail. The quality of these follow up charts depends on the experiences of construction managers. The user has to specify the construction activities and their interdependencies. Some of these interdependencies are of technological nature. The technological inter¬dependencies must be considered whereas other interdependencies, for example restrictions in the availability of resources, might be considered. Scheduling tools support construction managers with a wide range of func¬tionalities, e.g. the critical path method (CPM) or load balancing algorithms. However, these tools do not support the user specifying interdependencies in such a way that completeness and correctness of inter¬dependencies can be guaranteed. They document user input, and completeness and logical correctness of follow up charts can only be checked by inspection. This paper presents a modeling technique where the technological interdependencies between construction activities are treated as results of algorithms. Each construction activity is modeled individually and independently of other activities by describing its preconditions and its results. The interdependencies between activities are calculated from the activity-oriented specification based on relational algebra. A proposal for the order of activities is computed. However, rescheduling is still necessary to consider additional restrictions. The advantage in using the presented modeling technique is that logical correctness and completeness of the proposal can be guaranteed so that the quality of scheduling improves. A practical example is presented to illustrate how the modeling technique works and what results can be calculated to support work that is at present time specification work of construction managers. This paper describes the modeling technique in a context where specifically the use of this technique for construction scheduling is explained. Some more theoretical aspects have already been published (Huhnt, 2005). The approach presented in this paper is a new approach in the area of semi-automate approaches to generate construction schedules. Differences to existing approaches and advantages in contrast to existing approaches are described.