RUB arrow Bauwesen arrow Informatik arrow
alice02.jpg

Knowledge-Based Construction Scheduling

The construction schedule for the erection of a large structure can contain thousands of individual steps. The creation, extension and maintenance of a contruction schedule is, even today, mainly a manual task, relying on the experience of the planner. Because this experience is not readily accessible in a transparent and suitable manner, it cannot be used by planners to evaluate possible construction alternatives in a new project.

 

The individual tasks for the erection of a structural component, along with the necessary prerequisites and results, can be stored individually in a case-based data base. In the design of a new construction project, the proposed structure can be broken down into components that are comparable to those already stored in the data base. To this end, the data base has been augmented with comparision procedures to compare the similarity of structural components. Planners can therefore use the data base to find entries for each new structural component and retrieve the corresponding construction tasks.

 

Based on these tasks, a set of initial, overall construction schedules can be automatically generated. Suitable mathematical methods have been developed and implemented to generate such schedules, which may also  contain various construction alternatives. It must be noted, however, that a generated construction schedule is not always correct or complete. This can happen, for instance, if a structural component in a new project is not yet in the data base. Thus, a software with a graphical user interface und algorithms based on graph theory has been developed which can recognize and analyze construction alternatives and check each alternative with respect to completeness and correctness.

 

The proposed approach has been initally validated with simple examples. It have been proven that increasing the number of cases in the data bases as well as developing effective analytical methods greatly enhances the qualitiy of automatically generated construction schedules. In particular, the automatic generation of schedules is very advantageous in situations where unforeseen disruptions or other incidents require a change to existing schedules, usually on a short notice. The continued development of advanced visual presentation and interface components as well as the deployment of transparent selection and evaluation methods for design details can extend the concept of knowledge-based construction planning and schedulung to further research projects.

Project Data

Title: Knowledge-Based Modeling of Construction Alternatives

Type: DFG Research Project

Principal Investigators: Prof. Dr.-Ing. Markus König, Prof. Dr.-Ing. Karl Beucke (Bauhaus-University Weimar)

Researchers: Katja Shapir, M.Sc.

Project Partner:
Bauhaus-University Weimar
Informatik im Bauwesen
Coudraystr. 7
99423 Weimar

Collaboration Partner:
OBERMEYER
Project Management GmbH
Kantstraße 2
04275 Leipzig 

Contact:
Chair of Computing in Engineering
Bldg. IC, Room 6-59
Universitätsstraße 150
44780 Bochum

Links

funded by the

dfg.gif