Introduction In the previous article, “ Modular Data Analytics and Reporting in Facility Management – Five Modules,” we discussed utilizing five modules, including Asset Operation, Asset Maintenance, Safety, Real Estate, and Sustainability, from a modularized design perspective to encompass a holistic digital twin. Modular data analytics and report - ing enable performing quantification analysis for each focus area in a facility. The modular framework also provides a foundation for the end-user to obtain the most straightforward understanding of facility performance when a corresponding scoring system is established. This article disassembles each module’s health status into subcategories and associated parameters to demonstrate how a health scoring system can support digital twin implementation and facilitate performance track- ing in conjunction with facility operation and maintenance (O&M). Decision-Making Support System (DMSS) The quantifiable decision-making support system (DMSS) is usu - ally described as an information system whose purpose is to provide partial or full support for decision-making phases, including but not limited to intelligence, design, choice, implementation, and learning. In the AECO industry, a DMSS should be designed with the follow - ing two considerations. On one hand, it is common for current O&M practices to rely mostly on the facility operators’ knowledge and experience. According to the Operations and Maintenance Benchmarks issued by the Interna - tional Facility Management Association (IFMA) average building operation and maintenance costs jumped by 72 percent between 2007 and 2017. The cost increases were mainly due to greater expenses associated with maintenance staff and building deterioration. This dilemma illustrates the imbalance between growing O&M needs and a limited pool of seasoned O&M staff. Therefore, high O&M costs and low O&M efficiency should be expected consequences that most facilities need to plan for now or the near future. To overcome the asymmetry of the supply and demand relationship, it is critical to shift from the traditional experience-oriented subjective environment to a data-oriented objective environment. Introducing a standardized scoring system based on comprehensive data collection and analysis to replace human judgment based on personal experience will be the next step to accelerate the change. On the other hand, traditional Asset Management/Facility Management (AM/FM) data practices that were manual or semi-manual created in - formation silos that prevent even seasoned O&M staff from effective query or cross-reference as well as a timely response. For example, a large transportation agency has structural information, inspection re- ports, and asset information stored in discrete databases which makes SCORING SYSTEM FOR FACILITY HEALTH ASSESSMENT By Kai Yin, and Jiayi Yan with George Broadbent, Dr. Jeff Chen, and Dr. Eve Lin
data integration or collaboration almost impossible. Moreover, infor - mation silos, even when connected, cannot provide information about the appearance of the facility. The essence or the correlation between data cannot be represented. In another example, a traditional data reporting system could provide generic asset information, such as useful life, manufacturer, usage condi- tion, and so on. However, it cannot connect all these pieces to inform whether the usage condition (public, private) will contribute to a short - ened productive life. This will obviously impact the remaining value- and scheduled Preventive Maintenance (PM) routines, and in turn affect the AM/FM budget. The advent of Predictive Maintenance (PdM) is a solution through data collection via IoT, data analytics, AI algorithms, etc. The design of DMSS needs to be in line with the development of PdM, and all related data flows to ensure the data can be aggregated and interpreted to match the needs of the entire building/facility’s O&M assessment to the facility operator’s daily work and capability. Design Methodology of Scoring System to Support DMSS Modular Health Scoring System (MHSS) The core concept of the Modular Health Scoring System (MHSS) is to provide comprehensive parameter ratings that cover three main cat- egories – Present Status, Trend, and Risk – for each of the five modules within a digital twin. By assessing parameters or factors in all three categories, the MHSS can provide a thorough health score for each module and trace the problematic area based on the rating logic of the MHSS. Parameters under “Present Status” focus on presenting the most up-to-date operational performance based on the current value or real-time monitoring and analysis. Parameters in the “Trend” cat- egory indicate the directional trend and variations. Lastly, parameters under “Risk” are the correction factors of the overall health score. For example, in the Real Estate Module, the Current Occupancy Rate rep - resents the present operational performance. The Rental Growth Rate parameter under the “Trend” category shows the future pattern, as well Rent Overdue Amount and Rent Overdue Frequency are factors under the “Risk” category. The following example demonstrates the importance of including all three categories in the MHSS. Suppose the MHSS for the Asset Main - tenance Module is only based on the parameters under the “Present Status” category, such as Equipment Maintenance Rate, Inspection Time, and Maintenance Quantities, and all are under the tolerable per - formance thresholds. In that case, the facility manager won’t pay atten- tion to the operational performance. However, if the MHSS includes parameters under the “Trend” category, and all the assets are operating under normal conditions, some trending parameters might reveal po- tential maintenance hazards, such as the Returning Rate Growth. The comprehensive scoring system of “Present Status + Trend + Risk” can reflect potential problematic areas via different parameter combina - tions in time. The cumulated overall scores for every module can show the performance of each one. This allows the facility manager to locate the problem by looking up the scoring of individual parameters under different categories and modules and avoids guess work. Table 1 provides a sample of modular scoring parameters, categories, and category coefficients of the five modules. With the digital twin
37
october 2021 csengineermag.com
Made with FlippingBook Annual report