Warehouse and Distribution

Science

CHAPTERS 1 – 3
WAREHOUSE OPERATIONS
 Definition of a Warehouse

Warehouses are points in the supply chain where

product pauses and is touched.
Consumes both space and time (person-hours).
 Why have a warehouse?

To better match supply with customer demand

To consolidate product, reduce shipping cost
 Types of Warehouses

Retail Distribution Center (DC)

Service Parts DC
Catalog fulfillment or E-commerce DC
3PL Warehouse
Perishables Warehouse
 Warehouse Systems Determined by

Inventory characteristics
 Number of products, size, and turn rates
Throughput and service requirements
 Number of lines picked and orders shipped per day
Footprint of building & cost of equipment
Cost of labor
Material Flow
 Fluid Flow Model of Product Flow

 Insights from Fluid Dynamics: fluid flows faster in

narrower segments of pipe than wider segments
 Implies on average, an item will move more slowly through
the region with large inventory than it will through a region
with little inventory.

Figure 2.1
 General Warehouse Guidelines from Fluid Model

Keep the product moving

 Avoid starts and stops which require extra handling and
additional space
Avoid layouts that impeded smooth flow
Identify and resolve bottlenecks to flow
A product is generally
handled in smaller units
as it moves down the
supply chain.

A stock keeping unit, or

sku, is the smallest
physical unit of product
that is tracked by the
organization.

Upstream in the SC, flow

is in larger units, like
pallets.

Product is successively
broken down into
smaller units as it moves
downstream.

Figure 2.2
Popularity vs
Volume (flow)
Popularity (number of
times requested or
picks) is NOT highly
correlated with physical
volume (flow) of a sku.

Makes warehouse
design difficult because
it is hard to design
processed that work
well with skus that me
be any combination of
popular/unpopular nad
low-volume/high-
Figure 2.3
volume.
The most
popular 20% of
SKUs
This plot is the same
skus ranked from most
popular to least.

This is a typical plot

that shows that a small
fraction of the skus
account for most of the
activity. (80%/20%
rule)

It is easy to design
processes for popular
skus because they are
fairly predictable. Figure 2.4
The remaining
80% of the
SKUs
Conisder the skus in the
“long-tail” of the curve,
the 80% that are
requested infrequently.

Impossible to know
which of those skus will
be requested tomorrow.

They occupy most of the

space of the warehouse.

Have high safety stock.

Figure 2.4
 “Two Warehouses” in One

Essentially you have 2 warehouses:

The first “warehouse”
 Organized around a small set of predictably popular skus
 Easy to plan for
 Challenge is to manage flow efficiently
 Labor intensive
The second “warehouse”
 Predictable in “aggregate” only
 Harder to plan for
 Space intensive
 Challenge to hedge space and labor tradeoff
 Dedicated Storage

Each location is reserved for an assigned product

Simple to implement
Store more popular items in more convenient
locations
Workers “learn” the layout making picking more
efficient
Does not use space efficiently – on average half full
Inventory
Level
Idealized representation
of how inventory level
at a location changes
over time

Average inventory level

over time 50%

Figure 2.5
 Shared Storage

Assign a product to more than 1 location, when a location is

empty it is available for reassignment
More locations, less product in each location, so space is
recycled sooner
Better utilization of space than dedicated storage
Need Warehouse Management System (WMS) to direct workers
More time consuming to put away
Requires worker discipline to pick where directed not where
most convenient
May have more discrepancies between book and physical
inventory
 Theorem 2.1

When a sku is stored in k locations of equal size the

average space utilization is k/(k + 1).
Moving from 1 location to 2 locations, improves
utilization from 50% to 66%
Increasing the number of locations increases
utilization, but the improvement diminishes as k
increases.
Increasing the number of locations also increases the
management required.
 Storage In Practice

Typically the actual space utilization for shared

storage is slightly less than the value from Theorem
2.1
Shared storage is more often used for bulk storage
areas (i.e. pallets)
Dedicated storage for the most active pick area,
where the area is smaller and labor benefits matter
most.
Illustration of Shared & Dedicated Storage
Example of a Hybrid System
 Warehouse as a Queuing System
A queuing system is Theorem
   2.2 (Little’s Law)
where customers arrive For a queuing system in
and join a queue to await steady state that average
service by any of several length L of the queue
servers. After receiving equals the average arrival
service, the customers rate λ times the average
depart the system. waiting time W.
Fundamental result of
queuing theory is Little’s
Law.
 Little’s Law Example
Warehouse
  with 10,000
pallets that turn an average
of about 4 times per year.
L = 10,000 W = ¼ year
What
   labor rate is
necessary to support this?
Assuming one 8-hour shift
per day and about 250
working days per year,
there are about 2,000
working hours/year.
20 pallets/hour
 Super Club Music
 Super Club distributes recorded music
to retail stores. Such physical
distribution of music is, of course, a
dying enterprise, as it is being replaced
by distribution via the web.
 Super Club stores are divided in to
routes and each route is visited by a
delivery truck once a week on a regular
day.
 Each day they pick for about 8 routes,
which total about 100 stores.
 On average each store orders about 50
sku's and about 3 of each sku, for a total
of about 15,000 pieces per day.
 The warehouse knows these orders a
day in advance of picking and so can
plan its work in advance.
Unique challenges
A very, very few sku's will be very,
very popular and most sku's will
scarcely sell at all. In fact, it is not
unusual for 20 percent of the sku's to
sell one or fewer copies over a year.
Popularity is very fleeting; what is a
popular product now may be dead in
two weeks.
A large number of returns in the
music business
More information http
://www.warehouse-science.com/?b6a
9d0d0
 Super Club Music

Static shelving with

Pallet rack
overstock on top
 Super Club Music

Typical shelf 48 bays of flow rack

 Super Club Music

Packing orders into

Pick face of flow rack
cartons
Warehouse
Operations
Most of the expense in a
typical warehouse is
LABOR.

Most of the labor cost is

ORDER-PICKING.

Most of the time in

order-picking is in
TRAVEL.