Figure 2: An example of inconsistent data definition and ambiguous data structure in a Traditional Asset Management Data Dictionary
tions and attributes – a form of a data dictionary. However, once the dictionary is defined in the CMMS, it remains in the system after implementation. New classifications and attributes can be added but the established structure cannot be easily altered or refined. In this inflexible data dictionary environment, the cumulative data update during the O&M period makes the inconsistent and ambiguous data structure unavoidable (as illustrated in Figure 2). In order to prevent future ambiguity and inconsistent information, a complete asset data dictionary first needs to be cleaned and refined prior to uploading to a CMMS. This leads to the second commonly observed issue in tradi- tional asset management – scattered asset information. Outside the CMMS platform without a designated central platform, asset information is usually maintained and tracked by different in- dividuals from separate departments in various formats, including but not limited to BIM models, CAD files, Excel spreadsheets, paper binders, text files, and PDFs. All these different formats and languages need to be gathered before reconciliation and consolidation for these diverse data sources and information can be established. Then comes to the issue of reconciliation and consolidation of the data. Different departments have their own individual means to track their information in various spreadsheets. There are also numerous systems that might be input in different periods by different individuals. While there is no im- mediate solution to query all the existing classifications and attributes among these scattered spreadsheets, the go-to method of those data
collectors or managers is to come up with their own definition for their data and information at hand. As the example illustrated in Figure 2, three different asset definitions including a misspelled word are defined by three different individuals at three different time periods. Although the pumps in the example serve three different systems - hot water, chill water, and treatment water; they can be categorized and refined into one “Centrifugal Pump” without a typo, and have a consistent at- tribute naming convention, i.e. Manufacturer, since “Manufacturer,” “Maker,” and “Manufacturor” are essentially the same attribute (as shown in Figure 3). This is just one small example within a haystack of asset information. Furthermore, this dispersed and unsynchronized solution impedes teamwork and collaboration, limits data analytics support, and at the same time creates a lot of unnecessary rework throughout the larger organization. Instead of capitalizing on the abun- dance of the available data for intelligent operations and maintenance, the facility or data managers drain their time trying to find the most updated spreadsheet. A Solution that Bridge the Gaps – An FM-oriented DDMS Illumed from the current issues and users’ needs, a more fundamental and practical approach might be an FM Data Dictionary Management System (DDMS), which can support different stakeholders and is ac- cessible to various parties. ADDMS does not merely serve a dictionary or database. Its functionalities are specifically designed to address the commonly seen issues during PIM to AIM transition as well as during operations and maintenance. A well developed DDMS in this era should be designed as a web ap- plication focusing on an intelligent cloud-based solution to help reduce the effort required to manage asset data information. It should also maximize the quality of data, as well as streamline data flow interop- erability from delivery to the operational stage. It is a cloud-based, multi-tenant solution. Hence, there is zero infrastructure required to apply. It focuses on facility equipment and all the associated attributes that need to be collected and included for all the different pieces of equipment. It enables users to migrate from traditionally scattered spreadsheets and fixed CMMS solutions that do not support the under- standing of the data structure and performance to a fully functioning data management solution that provides a central and flexible platform
Figure 3: Confusion in CMMS query caused by data inconsistency
11
october 2020
csengineermag.com
Made with FlippingBook Annual report