This study proposes adata-driven approach for benchmarking energy efficiency of warehouse buildings.Our proposed approach provides an alternative to the limitation of existingbenchmarking approaches where a theoretical energy-efficient warehouse was usedas a reference. Our approach starts by defining the questions needed to capturethe characteristics of warehouses relating to energy consumption. Using an existingdata set of warehouse building containing various attributes, we first cluster theminto groups by their characteristics. The warehouses characteristics derivedfrom the cluster assignments along with their past annual energy consumptionare subsequently used to train a decision tree model. The decision tree providesa classification of what factors contribute to different levels of energyconsumption. Finally, we showed how a linear regression method is used to predictthe energy consumption based on relationships between strongly correlatedvariables, such as climate zone, number of working hours, and floor area. Withour proposed data-driven approach, decision makers can analyze and benchmarktheir warehouse building data, adopt best practices from existing solutions andmake better decisions when recommending high-impact energy reduction solutionsfor their warehouses.
Energy Efficiency, Sustainability, Data Analytics
Databases and Information Systems | Data Storage Systems | Software Engineering
Intelligent Systems and Decision Analytics
3rd PROLOG Project & Logistics 2017, May 11-12
City or Country
La Rochelle, France
LEE, Wee Leong; TAN, Kar Way; and LIM, Zui Young.
A data-driven approach for benchmarking energy efficiency of warehouse buildings. (2017). 3rd PROLOG Project & Logistics 2017, May 11-12. 1-8. Research Collection School Of Information Systems.
Available at: http://ink.library.smu.edu.sg/sis_research/3658
Copyright Owner and License
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.