No Detail Found!

CAD and Spatial Data Mining Framework for Petro-Chemical Pipeline Industry

Author(s)
M. Shaheen Department of Computer Science & Engineering, University of the Engineering & Technology, Lahore, Pakistan
M. Shahbaz Department of Computer Science & Engineering, University of the Engineering & Technology, Lahore, Pakistan
S. M. Ahsan Department of Computer Science & Engineering, University of the Engineering & Technology, Lahore, Pakistan
Khadim Hussain Department of Computer Science & Engineering, University of the Engineering & Technology, Lahore, Pakistan
Abstract
Decision making within industry is a crucial process that needs great deal of attention, breadth of knowledge and experience in multiple domains, intellectual abilities and comprehensive understanding of process flow. The ability of decision makers is constrained by the complexity of the matter, accuracy of data and the circumstances in which the decision is to be made. The planning & development panel in Petro-Chemical pipeline industry utilizes substantial resources for making decisions about pipeline infrastructure installations, emergency management and hazard prediction. The technical scenarios in pipeline industry are visualized on the grid of spatial data and analyzed under certain engineering codes and rules. The decision made after such an analysis is not immune to human interventional error, lack of expertise, complexity of computer aided design, over-utilization of time and proximity of error. This paper is aimed to build a framework for authentic interpretation of geo-spatial and cartographic data and to make decision for pipeline installation by extracting the hidden patterns and technical knowledge from this data. This framework will read and extract data from GIS and other cartographic systems and use data mining and machine learning techniques to identify influencing factors in decision making. The framework will focus on minimizing human interaction and try to automate the complex decision-making process by searching the solution in the complex hypothesis space. An optimized model of decision making is proposed which will minimize revenue loss, time consumption, consumers dissatisfaction and environmental hazards, which are apparently destined in existing decision-making mechanism. Historical data about several decisions in case of new installation as well as standardized rules from the industry will be used to train the system. New rules would evolve on the successful training which will be helpful to find solution.
Publication Details
Page(s) 55-60
DOI DOI not available
Published Journal: Journal of Pakistan Institute of Chemical Engineers, Volume: 39, Issue: 1, Year: 2011

Request the full text from PASTIC

0

Citations

0

Downloads

56

Views