Common LISP: the language
The rise of the expert company
The rise of the expert company
The Definition and Implementation of a Computer Programming Language Based on Constraints
The Definition and Implementation of a Computer Programming Language Based on Constraints
| Hi-index | 0.00 |
The project early design-stage indicative construction time estimate (Predicte) system is an expert system that is designed to provide indicative construction time estimates of concrete-framed, multistory building projects at the concept or early design stages when relatively little project information is available. Given a preliminary design concept for a multistory building, Predicte returns a listing of the major construction activities required to build the building, the activity duration times, and their start and end dates. In this chapter, we describe the construction time-estimation process and our automation of this process. We discuss the implementation of Predicte and how it was developed, tested" and deployed. We also discuss the maintenance of the system and its business payoff.The Predicte system was jointly developed by the Lend Lease Corporation of Sydney, Australia; the AI Applications Group of Digital Equipment Corporation; and the Digital Sydney AI Centre. Initial contact between Lend Lease and Digital began in early 1985, and formal development of Predicte occurred from July 1985 to October 1987.Today, Predicte is accessible from 12 different Lend Lease locations throughout Australia.Predicte is one of the first expert systems developed for the construction industry (see Feigenbaum, McCorduck, and Nii [1988] for a list of expert system applications in the construction industry). It is also the first expert system developed for construction time-estimation tasks. Our work in developing Predicte helped us define a new class of expert systems, which we call design-verification and design-evaluation systems. We begin this chapter by describing the time-estimation process and why the application of expert system technology matches well with the problem and with Lend Lease's business needs.