Artificial Intelligence Techniques for Generating Construction Project Plans

Abstract

Enhancements to the theory of network‐based planning techniques invented in the late 1950s, along with improved hardware and software environments, have resulted in elegant and powerful tools for planning today's construction projects. However, a fundamental limitation of these traditional planning techniques is that they are able to manipulate only the data generated by the planning process, not the knowledge used in generating the project plan. Artificial Intelligence (AI) techniques provide the means to generate plans, and to reason with, and provide explanations from, stored knowledge. However, the weak methods, employing little domain knowledge, originally used in AI for planning, proved inadequate for complex real‐life problems such as project planning. As a result, more recent research adopts the knowledge engineering methodology as an efficient approach for developing planning systems. This paper highlights the limitations of existing project planning tools, illustrates the power of AI techniques in the construction planning domain through a summary and critique of previous and current research in AI planning, and concludes with a suggested approach to developing AI‐based project planning techniques for construction.