To efficiently and effectively manage projects in the construction phase, project managers have to transfer information within the drawings to obtain a project schedule. However, practically, the result is usually arbitrary depending on the project scheduler's experiences. This research employs MD CAD objects as basic scheduling components to develop a model that can effectively generate a contract-driven schedule. Several principles are developed to obtain Object Relationship Matrix (ORM) and Object Similarity Matrix (OSM). Genetic algorithms are then used with ORM and OSM to generate the optimal sequence of constructing building components. Several constructive constraints, such as the resource limitation and the continuity of workflow, are also taken into consideration within the model development. And the Project Planning and Scheduling System (PPSS) is developed to enhance the practice of the proposed model. In addition, a case study is presented to examine the efficiency of PPSS. It is proved that the automated scheduling model is capable to generate an optimal contract-driven schedule, and provides multi-dimensional information for management.