Energy-aware Models for Warehousing Operations

Open Access
Author:
Anand, Vidyuth
Graduate Program:
Industrial Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
None
Committee Members:
  • Vittaldas V Prabhu, Thesis Advisor
Keywords:
  • Energy control
  • Queueing
  • Warehousing
  • Forklifts
Abstract:
There is a growing need in industries around the world to become more sustainable and reduce consumption of energy. Due to the rapid increase in demand of goods, there has been a rise in demand of logistics and operational services. This necessitates the need for a large number of warehouses and distribution centers to satisfy the demand. Industries across the country have given this a fair amount of consideration, as the amount of money invested in business logistics in the United States in 2012 was $1.33 trillion, at 8.5% of the GDP. Being a key player in the market, it is imperative that warehouses follow the same sustainable development model practiced in other industries. The U.S. Census Bureau reported approximately 11,000 warehouse and storage facilities operating in the country in 2008 with employment of close to 600,000 workers who earned annual wages of nearly $21 billion. Warehouses account for 8% of the total energy consumption of commercial buildings across the nation. Despite a higher level of warehouse automation in the market today, forklifts are still a critical component of warehousing activity, and contribute significantly to the consumption of energy in warehouses. The U.S. alone is responsible for shipping more than 100,000 units of forklifts around the world today, translated to more than $30 billion worth of forklifts being bought by warehousing companies. We serve to leverage the EC1 energy control policy model for manufacturing systems to forklift queueing models in warehouses as an effective energy metric. Specifically, warehouses are modeled as M/M/c queues of forklifts. The model is extended to general distribution queues and experimentation based on real-world data for different distributions is carried out to infer outcomes based on the model, following which a scope for future work is described.