Automating Schedule Review for Expressway Construction R. J. Dzeng 1 ; H. P. Tserng 2 ; and W. C. Wang 3 Abstract: An expressway project is often divided into subprojects with different tendering packages, and carried out by several general contractors that apply different scheduling practices. Each schedule may contain hundreds of activities, each of which is associated with multiple pay items that determine its earned monetary value. With such huge amount of information, the reviewer can only check a sample piece of information, and the quality of review highly depends on the reviewer’s experience and devotion. Automated schedule review provides a solution to reduce such problems encountered in the industry. This paper presents a module-based schedule generation and review model, which includes a predefined set of network modules, network builder assistant computer system that helps schedulers manage and reuse the modules to build a new schedule, and another computer system network review assistant (NRA) that helps reviewers review schedules. The NRA uses generalized rule forms to represent the schedule critique knowledge collected from the industry. When potential errors are found, the NRA adopts case-based reasoning to suggest possible correction based on similar cases. The evaluation conducted by the practitioners using real projects indicates that NRA reduces review time, and provides more accurate review on finding activities and related pay items not conforming to standards, and reminding users of important but often omitted activities. DOI: 10.1061/(ASCE)0733-9364(2005)131:1(127) CE Database subject headings: Highway construction; Scheduling; Construction management; Automation. Introduction Schedule administration for public agencies in charge of express- way construction includes work of specifying requirements for preconstruction schedules submitted by contractors, reviewing, and approving them before starting construction. It also includes the reviewing and monitoring of periodically updated schedules during construction. An expressway project is often divided into subprojects with different tendering packages because of many factors such as the volume of work and capital resources involved, work and finan- cial capability of the market, and balancing of excavated and refilled soil. The project is thus carried out by several general contractors who may use different activity names and levels of detail. Even after the division, each schedule may still contain hundreds of activities, each of which is associated with multiple pay items that determine its earned monetary value. With such a huge amount of nonstandardized information, the reviewer can only check for a sample piece of information. The review quality also depends highly on the reviewer’s experience and devotion. Automated schedule review provides a solution to reduce the problem by criticizing the submitted schedules, screening out po- tential errors, and possibly suggesting appropriate correction. It allows reviewers to be able to concentrate on reduced information that is more likely to contain errors rather than be submerged by information flood. It saves time for reviewers, and also ensures minimum review quality even though it does not necessarily guar- antee improved review quality. Research of automated schedule critique using artificial intel- ligence techniques has been found in the construction manage- ment literature. However, none has addressed the issues of sched- ule standardization, a primary assumption on the input schedule to be reviewed for an automated schedule critique system. An ad hoc standardization is a promising concept, but very often is hard to implement in practice because it requires professionals in- volved in changing their daily work practice. Dzeng (2000) de- veloped network builder assistant (NBA) to support the proposed “soft standardization” phase, a phase where standardization is en- forced by using encouraging tools and incentive instead of rigid legal documents. The NBA can help schedulers quickly build a preliminary schedule based on predefined network modules that comprise standardized activities and pay items. By using NBA, contractors may save scheduling time and effort, and also fulfill the standardization required by the client. Dzeng and Wang (2003) applied the modules to address the schedule integration issues among multiple contractors and the client’s multiple man- agement levels of typical highway projects. This paper describes our research on developing an automated schedule review system, called the network review assistant (NRA), for expressway construction projects from the perspective of a public agency. The NRA assumes that the given schedule to be critiqued applies standard activities. The schedule may be gen- erated by NBA, or other commercial scheduling software such as P3 (Primavera Systems Inc. 1999). The NRA can identify poten- tial schedule errors using rule-based reasoning and suggest pos- sible corrections using case-based reasoning. The paper is organized as follows. It first reviews the related literature on existing schedule review and generation systems. 1 Associate Professor, Dept. of Civil Engineering, National Chiao- Tung Univ., Taiwan 30050, ROC. 2 Associate Professor, Dept. of Civil Engineering, National Taiwan Univ., Taiwan 10600, ROC. 3 Associate Professor, Dept. of Civil Engineering, National Chiao- Tung Univ., Taiwan 30050, ROC. Note. Discussion open until June 1, 2005. Separate discussions must be submitted for individual papers. To extend the closing date by one month, a written request must be filed with the ASCE Managing Editor. The manuscript for this paper was submitted for review and possible publication on November 14, 2002; approved on February 24, 2004. This paper is part of the Journal of Construction Engineering and Manage- ment, Vol. 131, No. 1, January 1, 2005. ©ASCE, ISSN 0733-9364/2005/ 1-127–136/$25.00. JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT © ASCE / JANUARY 2005 / 127 J. Constr. Eng. Manage. 2005.131:127-136. Downloaded from ascelibrary.org by National Chiao Tung University on 04/26/14. Copyright ASCE. For personal use only; all rights reserved.