An information model for managing design changes in a collaborative multi-disciplinary design environment

The presence of incompatibility errors in construction technical documents is a major problem for the construction industry. An analysis of the production process of these documents reveals that managing design changes constitutes a main source for incompatibilities. More specifically, failure to propagate design changes among the design team is a principal cause of problems. The large amount of design data that is generated within a multi-disciplinary design environment makes this task very complex, especially when considering the fact that the involved design disciplines are, in most cases, separated both spatially and educationally. This thesis presents the development of a model that uses information technology to address the problem. The core concept of the model is a central database that functions as a repository of active building components. Each building component in this database not only carries its design data, but is also capable of recognizing the design disciplines that are affected by any change in these data and automatically send them messages. The active building components are able to perform this task because they are equipped with the necessary linking knowledge. Such knowledge is necessary to propagate the effect of a design change by one discipline on other disciplines involved in the design of the same building. The linking knowledge is acquired from the designers and is implemented in the form of rules. A management database is also developed as part of the model's central database. This management database makes the model easily adaptable to any building configuration, an essential requirement due to the uniqueness of every building project. It also provides the model with the capability for not only tracking past design changes but also planning, and scheduling future design changes as well. The model has been successfully implemented on a client-server network environment and validated using both hypothetical and real design cases