Geotechnical engineering is something of a black art, relying on subjective judgment and feel as well as engineering principles. Consequently there is immense scope for using the benefits of artificial intelligence for geotechnical engineering applications where there can be a considerable variability of observed information and possible deductions. This paper describes an artificial intelligence expert system to simulate geotechnical laboratory tests. This is twofold in purpose: firstly to create an interactive knowledge-based expert system to predict the likely results of a geotechnical laboratory test using a physical description of the soil in question; and secondly to simulate the execution of the laboratory test as an educational computer-aided learning exercise to link geotechnical theory to practice. The complete expert system was conceived as a pictorial application with considerable use of the graphical elements of a computer interface. This application of artificial intelligence requires a knowledge base of test results and related data, besides an inference mechanism that can predict the geotechnical parameters together with their appropriate precision.

The expert system has been built using HyperCard software, running on an Apple Macintosh computer. HyperCard uses a non-traditional concept in programming to enable a highly graphical interface to be easily constructed. The software is object-orientated for the graphical interface elements as well as for the programmed text. An introduction to the HyperCard software is presented and the expert system is explained in both its predictive mode and its tutorial mode.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £60 +P&P)