سیستم پشتیبانی تصمیم (DSS) مبتنی بر انبار داده برای مدیریت زیرساخت فاضلاب

ABSTRACT A data warehouse-based decision support system for sewer infrastructure management

Currently, most municipal agencies are facing challenges regarding the deterioration of underground infrastructures due to the lack of available funds and experienced workers for renovation. According to a recent report [1], underground infrastructures in the United States are now graded D— (on a scale of A to F and I), and up to $255 billion of funding is potentially required to keep the infrastructures running smoothly for the next five years. This means that most sewer systems are approaching a state of near collapse, and undetected defects are becoming increasingly more severe. In response, government agencies are paying more attention to the implementation of infrastructure management systems, which will enable agencies to manage and improve their sewer systems more efficiently and cost-effectively.

An infrastructure management system refers to a systematic approach that optimizes the maintenance, operation, preservation and improvement of physical assets throughout their service lives at minimum cost [2]. Such systems generally include the functions of inventory of assets, condition assessment, rehabilitation and replacement, prioritization, and investment decisions. The actual structures may differ for each agency, but the inherent concepts are the same. The series of aforementioned processes are linked to one another in stages through a shared flow of pertinent data. Thus, one key to a successful infrastructure management system is how well agencies organize, analyze, and utilize an enormous amount of data to ultimately achieve desirable service delivery.

Since data obtained by the business processes of infrastructure commonly provide predictable clues about the future performances of systems that guide long-term investment plans for the inspection and renewal of structures, this data should be safely stored while still being easily accessible for further analysis. The concept of a data warehouse well satisfies these requirements. As explained by Chau et al. [3], the objective of constructing a data warehouse is to provide a system that allows proper data to reach the right end user at just the right time. Thus, the primary purpose of the implementation of data warehousing for sewer infrastructure systems is to provide relevant and timely information to the infrastructure managers in an easily understood format so that customer service decisions can be made more efficiently and effectively. To managers of infrastructures, the phrase “data warehousing” does not merely indicate an efficient tool for data integration, but also implies the materialized format of a visualized, real-time management tool encompassing project design, rehabilitation and improvements.

The expected roles of data warehousing in sewer infrastructure management systems would include reasonable allocation of resources to meet required services, projection of the overall state of sewer lines, and identification of potential risks within sewer systems. To assist with these performances, this study presents a prototype Decision Support System (DSS) using data warehousing. Its core functions are to automatically determine the most appropriate options for inspection and renewal of sewer pipelines under specific conditions and to simultaneously estimate associated costs. Additionally, it provides utility managers with an analysis platform for comprehensive understanding of sewer infrastructure systems from various managerial points of view.

There have been recent efforts toward the development of decision support systems for renewal of pipelines in underground infrastructures, including AUTOCOP, PipeADDIN, H2ONET designer, CARE-S, CARE-W, and SIROCO. AUTOCOP was an evaluation program that used the Analytical Hierarchy Process (AHP) and a group decision model to process the various sets of criteria and sub-criteria of the decision problem [4]. AUTOCOP was applied to the selection of suitable renewal methods through the comparison of four alternatives: open-cut, micro tunneling, pipe bursting, and cured in place pipes.

PipeADDIN was mainly composed of three models: the pipe load model, the pipe deterioration model, and the pipe break model [5]. PipeADDIN computed the loads and stresses on a water main, estimated corrosion-induced deteriorations, and assessed vulnerability based on critical safety factors in order to prioritize pipes for future renewal.

H2ONET designer was an AutoCAD-based graphical tool utilizing a variation of the Genetic Algorithm to determine which alternative had the lowest cost while still satisfying hydraulic requirements for networks [6]. CARE-S and CARE-W were developed to assist governmental utilities in making better financial decisions by establishing multi decision criteria including technical and economic conditions [7,8]. While CARE-W was mainly intended for larger water companies with sufficient data, the primary objective of the SIROCO initiative was the provision of a water pipe prioritization tool for small- and medium sized water distribution companies with limited data [9]. SIROCO utilized the Weibull model and Monte Carlo simulation to estimate probabilities of pipe breaks during any given period of time.

Although these previous efforts provided advanced decision support functions for underground infrastructure management, data warehousing has been rare. Compared with other DSSs, a data warehouse-based DSS is relatively simple and easy to use. This study presents a user-friendly, computerized DSS for sewer pipe management utilizing the input of expert opinions and specifications of inspection and renewal methods within the framework of the DSS. This feature leads to enhanced applicability of the system under limited utility data.

The primary objective of this research was to develop a condensed but effective sewer infrastructure management system in the form of a data warehouse. The data warehouse was combined with decision support modules to allow the intelligent processing of any given sewer management system. Accordingly, an agency could comprehensively understand its current status, estimate budgetary requirements, and easily produce ad hoc reports about infrastructures with the appropriate level of required detail; this would lead to the reduction of inherent risks over the entire infrastructure systems through the reasonable allocation of resources.