Discipline: Technology and Engineering
Subcategory: Civil/Mechanical/Manufacturing Engineering
Binita Guchait Saha - Texas Southern University
Co-Author(s): Yachi Wanyan, Texas Southern University
Soil Classification is one of the most basic procedures used to identify soil properties. Commonly used soil classification systems include U.S. Department of Agriculture (USDA) textural classification (graphical), American Association of State Highway and Transportation of Officials (AASHTO) method (tabulated) and Unified Soil Classification System (USCS) (decision tree). Every Civil Engineering (CE) student needs to learn different ways to classify soil in order to design, analyze and construct infrastructures. Knowledge Based Expert System (KBES) as one of the many Artificial Intelligence (AI) applications used in CE is a good fit to automate the procedure based on the following hypothesis: 1) KBES works well on narrow and specific knowledge domains; 2) The rapid prototyping abilities of KBES makes it a perfect tool for soil classification; 3) KBES can handle different forms of information such as graphic, tabulated or general decision tree to provide decision-making support. First, simple if-then rules were used to reproduce the logic and formulas for each classification system. These rules were then entered in Microsoft excel. To test the rules, soil classification laboratory data were used as control to validate the outcome. Once these little decision-making blocks were functional, three independent classification blocks for the aforementioned soil classification systems were developed. These blocks were later ‘assembled’ together with the help from the mentor by using the CORVID developmental package. This automated tool provides quick and reliable soil classification results, which is especially helpful to users who are newly exposed to soil classification. More importantly, this ‘Lego-building’ approach provides a perfect hands-on learning project for undergraduate students. With computer based soil classification program developed, our next step is to develop a smart phone app and a web-based version for users.
Reference: The United States Soil Classification System and Its Application In Arizina. (n.d.). Retrieved August 02, 2016. Das, Braja M. Principles of Geotechnical Engineering. Boston: PWS, 1998. Expert System. (n.d.). Retrieved August 02, 2016. Americal Association of State Highway and Transportation Officials (1982). AASHTO Materials, Part I, Specifications, Washington, D.C. AASHTO Soil Classification System. (n.d.). Retrieved August 03, 2016. Soil Classification. (2010). Retrieved August 02, 2016.
Funder Acknowledgement(s): This study was supported by the National Science Foundation (NSF) HBCU-UP award HRD 1533569. The undergraduate researcher would like to express appreciation to the project’s faculty advisor, Yachi Wanyan, for mentorship and guidance.
Faculty Advisor: Yachi Wanyan, wanyany@tsu.edu
Role: In this research project, my contributions go purely to the soil classification test in the laboratory and from there to create three different excel sheet using “if – then” rules for each of the soil classification.