Achieving Cost-Effective Supply Chain
Agility via Regression and Optimization
Mariah Jeffery
Renee Butler, PhD
Department of Industrial Engineering and Management
Systems
mcmurran@mail.ucf.edu
Supply chain agility has received a great deal of recognition in recent
literature as organizations seek to gain a competitive advantage by determining
successful and cost effective agile strategies. We present an approach to
quantify and optimize supply chain agility (the ability to effectively respond
to supply and demand uncertainty) and its cost based on data mining, regression
modeling, and operations research. Data is collected related to on-time
delivery performance, supply chain settings, and other factors at the time of
order delivery. Ordinary least squares and logistic regression are used with
delivery performance as a response to develop an objective function, which is
optimized stochastically in two planning horizons: strategic and recourse.
Outcomes include the determination of inventory levels and transportation
speeds that result in the most cost effective agility level for specific
products, as well as insight into the effects of uncontrollable factors such as
forecast accuracy, customer lead times, and demand variability on delivery
performance. These insights can be applied to the determination of price
incentives to offer to customers for purchasing on contract and aid
organizations in determining where to focus improvement efforts.
We
present preliminary results from a case study application of the methodology in
the semiconductor industry including equations for agility and cost. Expected
outcomes and contributions as well as future research are also provided.