| dc.description.abstract | Collaboration between members of the supply chain is advocated for improving
efficiency and reducing the cost of delivering the products. Mechanisms to coordinate
decisions and share information between members of the supply chain have become
feasible options due to recent advances made in the area of Information Technology.
Considering a single-supplier, multiple-distributor setting, we show how the use of
relevant information shared through collaboration aids the decision making process of
the supplier. We incorporate information flow related to demand distribution
parameters and the realized demand at the distributors' end along with the orders
placed by the distributors in estimating the demand in each period of the horizon for
the supplier.
Based on the extent of information shared by the distributors with the supplier
through collaboration, following levels of collaboration are defined for the study.
a) Collaboration Level 0: Supplier takes replenishment decisions based on the
knowledge derived from the distributors' order history.
b) Collaboration Level 1: Supplier takes replenishment decisions based on the
knowledge derived from the distributors' demand distribution parameters for each
period of the horizon.
c) Collaboration Level 2: Supplier takes replenishment decisions based on the
knowledge derived from the realized demand at the distributors' end along with
the distributors' demand distribution parameters for each period of the horizon
We show that depending upon the level of collaboration; the additional information
can be used to readjust the inventory policy parameters at the supplier through our
inventory policy model. The decision problem at the supplier resembles a finite
horizon, non-stationary stochastic multi-echelon inventory policy problem. We
develop a stochastic dynamic programming solution procedure to solve the decision
problem of the supplier and compute the order up-to levels for each period of the
horizon minimizing the supplier's average inventory holding and stock-out cost over
the entire horizon. Based on ow experimentation, the demand range approximation
considered captures real-world complexities with less than 0.3% deviation from best
case analysis and less than 0.05% increase in average cost. The computation effort
reduces by linear level with increase in the magnitude of demand range.
We rigorously measure the value of information sharing through numerical analysis
based on an experiment design developed to represent various operating
environments. This can serve as a decision support tool for firms wishing to
collaborate with their supply chain partners.
The main results are:
The cost reduction can be as high as 70% for the situation when the supplier
identifies that the mean demand at the distributors' end varies across different
periods in a range which is 100% of the mean value.
If the distributors face high variability in demand in a period, then the extent of
benefit of information sharing will reduce. For a case when the mein demand
varies by 100% over the mean value across different periods, the cost saving will
only be up-to 40% (coefficient of variation is 30%) instead of being up-to 70%
(coefficient of variation is 10%).
If the distributors share information related to the realized end consumer demand
with the supplier in each period within their replenishment interval along with the
demand parameters, the supplier could additionally reduce costs up-to 47%.
The total cost saving at the supplier could rise up-to 82%, if information related to
the uncertainty of distributors' demand is shared along with the information
related to end consumer demand.
As more information related to distributors' demand and operating environment is
made available to the supplier, it would be able to serve the distributors in a better
way. The number of units short (stocked-out) at the supplier will reduce
marginally (by up-to 0.2%).
We also have taken an organization's supply chain, as an example to demonstrate the
need for collaboration between members of the supply chain. Using simulation model,
we compare the supplier's performance under the present practices within the supply
chain with that under a proposed inventory policy for the supplier based on the
knowledge derived from distributors' order history. This serves as a benchmark for
supplier's performance for future savings possible due to collaboration with the
distributors. We show that under the proposed policy the supplier can benefit between
40 - 75% in terms of Percentage Cost Benefit while maintaining almost the same
levels of customer service (Order Fill Rate and Item Fill Rate). | en |
