All over the world, local authorities are moving towards managing and storing urban data in digital form. But the data storage devices used are heterogeneous and typically include relational database management systems (DBMS), GIS, and CAD files. As a result, data are present in different locations on different platforms and under different schemas. This poses a problem for software applications meant to support decision-making in urban design that require input from more than one data source.

This dissertation demonstrates how data warehousing—combined with object oriented data modeling—is able to provide a general solution for this problem. Data warehousing is a technique initially developed for business applications, but is equally useful for urban design. The data warehouse constitutes a communication layer between the urban design applications and data sources. It makes the data available through a unified interface that hides the sources themselves and represents that data in terms of a general-purpose, preferably object-oriented, model. The data model is developed based on an analysis of the data needs of typical applications that support urban design and planning. Key components of the proposed warehouse architecture are Data Access Objects communicating with, data extractors able to handle queries expressed in terms of the data model, and solvers that detect contradictions in data and are able to perform data cleaning. This dissertation also describes a first prototype implementation of the data model and the data warehouse.

The test case used in this research to validate the object model and data warehouse architecture is the city of Makkah in Saudi Arabia, which faces significant urban design and planning challenges in connection with the annual pilgrimage that brings millions of visitors to the city. This research is sponsored by The Custodian of the Two Holy Mosques Institute for Hajj Research in Makkah, Saudi Arabia.