Inventory: Managing

Economies of Scale

Arijit Mitra
I know you have learnt a lot about
inventory in OM – 2!

• I have, therefore, decided NOT to discuss the basic models

like EOQ and POQ for more than 5-10 min!

• I will rather start this topic by asking you two simple

questions about inventory

• How do y ou segment your products /SKUs so th at y ou can handl e

them strategically?

• How to handle the uncer tai nty while handling inventor y?

Product Segmentation in Supply Chain

How can I segment my products?

• Physical characteristics (value, size, weight, density, etc.)

• Demand characteristics (sales volume, volatility, sales

duration, etc.) – Most Common

• Supply characteristics (availability, location, reliability, etc.)

Analysis of the Dry Goods SKUs of a local
grocery store.......

• ~ 8000 SKUs

• A total of 1.156M SKUs sold in 1 year

• Number of units sold per SKU

• Mean 144

• Median 72

• Mode 0

• Std Dev 355

Some Questions which may come to the
analyzer's mind..............

Top Sellers
• What are the Biggest Sellers?
1. EVAPORATED MILK 12 OZ
2. BATHROOM TISSUE
3. BOTTLED WATER 1 GALLON
4. MAC’N CHEESE
5. CANNED WHITE TUNA

• Biggest Sales Day? – 3 Big US Holidays when sales were 2

times as compared to normal sales days!

• How are products distributed in terms of sales volume?

Uniform? Normal? Other?
Potential Product Distributions

Lognormal Distribution

Normal Distribution Power Distribution

Uniform Distribution
Potential Distributions for our SKUs

Following each
distribution, we
can plot the
cumulative % of
sales volume
against the % of
the SKUs involved
in the sales
 1.2

0.8
y = 1.1245x0.3784
% Sales Volume

0.6 R² = 0.9717

0.4
Model Fit
0.2
for Our
0
0 0.2 0.4 0.6 0.8 1
SKUs
% SKUs

• An Example of Power Law, y = a*xk

• Critical Few Versus Trivial many

• Is this distribution unique for only our SKUs?

What theory says?

Po w e r l a w i s E x c e p t i o n a l l y c o m m o n i n p h y s i c a l a n d s o c i a l s y s t e m s

• S e ve r i t y o f h u r r i c a n es a n d e a r t h q u a ke s

• I n c o m e w i t h in a p o p u l a t i o n ( Pa r e t o ’s L a w )

• V i s i t s t o w e b s i t es ( N i e l se n’s L a w ) & b l o g s

• F r e q u e nc y o f w o r d s i n a n y l a n g u a g e ( Z i p f ’s L a w )

• F r e q u e nc y o f d i g i t s w i t h i n t a b l e s ( B e n fo r d ’s L a w )

• F r e q u e nc y o f a u t h o r s c i t a t i o n s i n l i t e ra t u r e ( L o t ka’s L a w )

• A n i m a l s ’ m e t a b o l i c ra t e s w i t h r e s p e c t t o m a s s ( K l e i b e r ’s L a w )

• Profitability of customers & products

• D i s t r i b u t i o n o f vo l u m e o n t ra f f i c l a n e s

• Questions from students in a class

So, we use The Pareto Rule while
analyzing the inventory of SKUs

An d .......t h e r efo re, w e g o fo r t h e w e l l -k n o wn t e ch n i q u e A BC a n a l y s i s

w h i ch i s t h e r e s u l t w e g e t w h e n w e u s e t h e Pa r eto Ru l e .

A problem:

Accor d i n g to t he T h u mb r u le use d i n A BC a n a ly s is (w h i ch i s n o t h i n g
b u t t he Pareto r u le) , on l y 20% of t he SKUs g ives 80% of t he t ot al
s a les vo l ume ( A cate gor y Item s) a n d 50% o f t he S KUs g i ves us 95% of
t he to t a l s a les vo lu me (A a n d B item s t o get her) . If yo ur ware h ouse
i nve ntor y fo l l o w ex a ctl y t h is p attern , w h at mo del w i ll f it the
d i s tr ib u t i o n a cco r d i ng t o t he p ower l aw (f i n d t he va l ues of a a n d k
w h e re t h e g e n e ra l e q u a t i o n i s y = a * x k ) ?
Y = a*X^k

If X = 0.2, y = 0.8

And if, X = 0.5, y = 0.95

Ln(y) = Ln(a) + k*Ln(X)

-0.22314= Ln(a) + k*( -1.60944)

-0.05129= Ln(a) + k*( -0.69315)

a = 1.081 and k = 0.1875

Remember while performing ABC
analysis...

The basis of ABC may be different according to the strategic

focus of the company, the function which applies the technique,
etc.

• The value or price

• The volume of sales (number, revenue or profit)

• The velocity of the sales

• The volatility of demand

• And so on...
A quick recap of ABC analysis and
model fitting for an SKU

Datasheet gives us the information about the cost /unit and

annual usage of different component parts of a mechanical
equipment. Do the ABC analysis and fit the distribution of these
items into a model according to the power law (y=a*x k ).

Let us see the datasheet

• Af ter understanding this apparently simple (b ut most
important) concept, w e w ill start managing the in ve ntor y in
Supply Chain Two Conflicting objectives, The
Inventory Holding Cost and Ordering
• The basic equation in this regard is termed as th e “ Total
(Inventory) Cost” Equation Cost

• TC = Product Cost + Inventory Holding Cost + Ordering cost

• The total of different cost must be minimum

• Trade -off imbedded into that ; it is used to Operat ionalize

the Trade-off Theory

• Minimiz ation of TC Equation gi ves you the concept of

“Total Least Cost” (TLC) and the corresponding no. of
ordered quantit y is termed as EOQ or POQ (as the case ma y
be!)

• We will start with these basic models and wi ll move

towards more advanced models
Different Types of Inventory

Accounting Categories Functional Categories

• Raw Materials Inventor y • Cycle Stock

• Work-i n-Process (WIP) Inventor y • Pipeline Inventor y (PI)

• Components, Semi -f inished Goods • Safety Stock

• Finished Goods

• Tools and Equipment

Inventory Policy & Replenishment
Models

The primary objective is to find an optimal policy for managing

inventory. The policy is the Guidelines concerning

• What to purchase or manufacture

• When to take action

• In what quantity

• Consideration of a Replenishment Model / Inventory Policy –

Total Relevant Cost
Replenishment Model Assumptions
Inventory Chart & Basic EOQ Model
Model Assumptions for EOQ:

• Demand – Constant, known and

Continuous
• Lead time – Instantaneous (0)
• Independent Items
• Continuous Review
• No discounts
• No excess Demand
• No perishability
• Number of items – Single
• Total amount ordered is received
Economic Order Quantity (EOQ)

Time between replenishments

𝟐𝑫𝑺
EOQ or Q* = 𝟐𝑺
𝒉𝑪 T* = Q* / D =
𝑫𝒉𝑪
Wh e r e,
Total Cost = Product Cost + Ordering Cost + Inv. Holding Cost
D = An n u a l d e m a n d o f t h e
= CD + (D/Q*)S + (hCQ*/2)
P r o d u ct
Substituting the value of Q* from the EOQ Equation, we get
S = F i xe d co s t i n cu r r e d p e r
order Total Cost = CD + 𝟐𝑺𝒉𝑪𝑫
C = Cost per unit Total Relevant Cost = 𝟐𝑺𝒉𝑪𝑫
h = H o l d i n g co s t p e r ye a r a s
a f ra cti o n o f p r o d u ct co s t
EOQ Replenishment Models: Extensions

• EOQ with Instantaneous Lead time (The Basic Model)

• EOQ with Non-Instantaneous Lead Time

• Quantity Discounts
• A l l Q t y. d i s c o un t s

• M a r g i n al Q t y. D i s c o u nt s Extensions of EOQ – The

Assumptions change
• Tra d e P r o mo ti o n a n d O n e - t im e B u y

• EPQ or POQ

• EOQ with Planned Backorders

EOQ Extensions: Aggregating multiple
products in a single order

• Lots are orders and delivered independently (Simple EOQ

applied)

• Lots are ordered & delivered jointly (Simple Aggregation)

• Aggregation with Capacity Constraint

• Lots are ordered and delivered jointly for a selected subset

of the products (Tailored Aggregation)
 Uncertainty
in Inventory
Management
- Basics
Comparison of Inventory Charts for
Deterministic Demand and Stochastic Demand

Demand

Order

Receive

Inventory
Position

Stockouts
 Handling uncertainty in Supply Chain

• Managing to the “mean” or “average” is rarely sufficient!

• We need to handle the Variability i.e. the spread in the

data

• The variability or uncertainty is generally handled by

assuming a probability distribution

• Empirical Distribution

• Theoretical Distribution:

• Uniform Distribution (Discrete and Continuous)

• Normal Distribution

• Poisson’s Distribution
Three Critical Questions to Answer

1.What is the probability of Stocking Out for a given

inventory level?

2. How much inventory is required to meet a target

service level?

3.What are the expected values of units sold and units

short for a target inventory or service level?
A motivating Problem to answer these three critical
questions when demand follows different distributions

Assume that you are running a seafood outlet at Chilika, and your outlet,
“Lobsterly Yours” is well known by its lobsters. Tourists and locals come to
your shop to try delicious lobster that your shop offers. Harish, your Head
Cook and Manager must decide how many lobsters to buy from the local
fishermen everyday.

Your shop is committed to serve only the freshest lobsters, caught and
cooked on the same day. Therefore, he only has one opportunity a day to
buy lobsters (to the fishermen in the early morning). And he cannot store
any lobsters to be cooked the day after. Could you help Harish decide how
many lobsters to buy every morning?

Your shop has the daily demand data (in units of lobsters) for the last year.
Example of Empirical Distribution

Lobster Demand No. of Days

0 0
1. How many lobster Harish needs to reduce the
1 5 probability of stocking out to less than 5%? (Ans – 21)
2 7
3 10
4 14
2. What is the probability of stocking out if Harish
5 23 buys 15 lobsters in the morning? [P(SO) = 0.2055]
6 32
7 30
8 30 3. What is the expected units sold & what is the
9 25 expected units short if Harish buys 15 lobsters in the
10 24
morning?
11 23
12 20
13 19
14 16
15 12
16 13
17 11
18 11
19 8
20 7
21 8
22 6
23 6
24 4
25 1
Example of Discrete Uniform Distribution

Let us solve the same problem when 1. How many lobster Harish needs to
demand varies with a uniform distribution reduce the probability of stocking out to
having a min. demand of 6 lobsters per less than 10%?
day to a max. demand of 14 lobsters per
day; it is being equally likely that the 2. What is the probability of stocking out if
demand of lobster at a day can take any Harish buys 10 lobsters in the morning?
integer value (it is discrete, that is why!)
between 6 to 14. 3. What is the expected units sold & what
is the expected units short if Harish buys
10 lobsters in the morning?
Again we need to answer the same three
critical questions.
Poisson’s Distribution: The Slow-
moving Stocks

• U se d i n th o se s i tu a t i on s w h e n w e are i n te re s te d i n me a su ri n g h o w
man y ti me s a c e rt ai n e v e n t occu r s i n a sp e ci fi c t i m e i n te rv al o r i n a
sp e c i fi c l e n g th o r are a .

• I t may b e u se d to an al y z e Sl ow - M ov i n g i te m se l l i n g e v e n ts

• Pro b ab i l i ty th at th e e v e n t X w i l l occu r for x ti me , i . e .

𝒆 −𝝀 𝝀 𝒙
P(X=x) = where,
𝒙!

λ = averag e no. of succ esses within a given

region /duration

• Pro b ab i l i ty th at th e e v e n t X w i l l occu r for x ti me o r l e ss, i . e .

𝒆 −𝝀 𝝀 𝒊
P(X≤x) = σ𝒙𝒊=𝟎
𝒊!
Handling uncertainty with Poisson’s Distribution: An
example

At your “Quick Munch” bagel shop, you have some cakes

that you sell that are slow movers. The chocolate cake,
for example, only sells on average 2.5 cakes per week.
You have noticed that the demand follows a Poisson
distribution.

• What is the probability that you will sell more than 3

cakes in a week? [Ans: 0.242]

• What is the probability that you will sell no cakes in a

week? [Ans: 0.082]

• Suppose you have 5 cakes made ready to sell. What is

the probability that you will sell all your cakes in one
week? [ Ans: 0.109]
Handling uncertainty with Normal
Distribution – The Single Period
Inventory Model

Classic example is “The Newsvendor Model” for the single-

period or perishable inventory

The premise is simple. Imagine vendors, selling newspapers on

the street. Each morning, they have one chance to buy
newspapers in bulk from the printer. How many copies of
today’s paper should a vendor stock, knowing that unmet
demand means loss of sales and unsold copies end up
worthless?
 Newsvendor Problem: An example

Let’s say the vendor buys papers in bulk for $0.50 each and
sells them for $2.50. Every paper sold generates $2.00 in
prof it, and at the end of the day unsold papers are
discarded for a $0.50 loss.

• Do you think the vendor should keep some extra

Newspaper than the expected demand?

• Do think the vendor should keep papers less than that of

the expected demand?

• How many more or less than the expected demand?

 Newsvendor Problem: An example (Contd.)

To find out the number of Newspaper that the vendor will buy,
we need to understand the nature of the demand

Assume that the demand of the Newspaper is distributed

normally, and the expected demand (D μ ) = 100 with a standard
deviation (σ D ) of 15 (both measured from the historical demand)

• What is the cost of an Overstocking (1 more unit than D μ )?

• What is the cost of an Understocking (1 less unit than D μ )?

 Newsvendor Problem: An example (Contd.)

Cost of Overstocking (C o ) = $0.50 (the scrap cost)

Co st o f Understo c k ing (C u ) = ($2.50 - $0.50) = $2.00 (the c o st o f lo st

sales)

Then The ideal po int o n the demand distributio n o r the Critic ality ratio
(Pro bability value) is c alc ulated as,

P = C u / (C o + C u )

In this c ase, P = $2.00 / ($2.00 + $0.50) = 0.8

The no . o f New spaper to be bo ught, X = D μ + z P * σ D

F ro m Table, z 0.8 = 0.7995 ans so ,

X = 100 + (0.799 5 * 1 5) ~ 113 [make sure that you are rounding off the
result to the next integer!}
ROP, Safety Stock
& Inventory Review
Policies

Arijit Mitra
Most Frequently used Replenishment Policies
1. EOQ Policy (with lead time L) – deterministic demand

• Order Q* when Inventory Position = DL

• Order Q* every T* time periods

2. Single Period Models – variable demand

• Order Q* at start of period where P[x≤Q] = Critical Ratio

3. Base Stock Policy – one-for-one replenishment

• Order what was demanded when it was demanded

4. Continuous Review Policy (s,Q) - event based

• Order Q* when Inventory Position ≤ s

5. Periodic Review Policy (R,S) – time based

• Order up to S units every R time periods

Re-order Point For Deterministic Demand
• EOQ policies with lead time L

• ROP = D*L = μDL

• Example:
• Annual demand = 52000 units
• Week per year = 52
• Lead time = 1 week
• ROP = D*L = μDL = 52000*(1 week / 52 week) = 1000
• Order Q* when the stock reaches 1000 units

• Remember the unit (week, days etc.)

Reorder Point for Stochastic Demand
• ROP = D*L + Safety Stock = μDL + z*σDL Safety Stock (SS)

• Example:

• Annual demand, D ~ N (52000, 5000)

• Lead time = 1 week

• μDL = 52000*(1 week / 52 weeks) = 1000 units

𝟏
• σDL = 5000* = 693.37 units (Remember the units!)
𝟓𝟐

• ROP = 1000 + z*693.37

• The z value depends on the Cycle Service Level (CSL) that the management wants

• Suppose, the CSL is 95% i.e. the probability of Stock out is 5%

Safety Stock (SS)
depends on CSL!
• z = normsinv(0.95) = 1.644

• ROP = 1000 + 1.644*693.37 = 1000 + 1140.5 = 2140.5 units; Order Q* when the
stock level reaches 2141 units
Quick Aside on Converting Times

• Ds = Demand over short time period (e.g. weeks, days etc.)

• DL = Demand over long time period (e.g. quarter, year etc.)

• n = No. of short periods within a long

• Mean of DS = (Mean of DL) / n and Mean of DL = (Mean of DS )* n

• SD of DS = (SD of DL ) /√n and SD of DL = (SD of DS )*√n

• Again, remember the units!

Base Stock Policy
• The maximum storage capacity is S*

• One-for-One Order Policy

• According to this policy, orders are placed, as and when the

inventory level drops from S*, the base stock level (or Inventory
Position) due to a demand

• Finding the S* (S* is constant) L

• S* = ROP = μDL + z*σDL

• z is determined either by management or the organization

follows a critical ratio approach where S*

• CR = Cu / (Cu + CO)
Base Stock Policy: An example
• Set a base stock policy for an item
• Daily Demand, D ~ N(200, 30)
• Lead Time is 3 days
• CO = $10 per day and CU = $25 per day

• Solution:
• μDL = 200*3 = 600 units
• σDL = 30*√3 = 52 units
• CR = 25 / (10 + 25) = 0.714 and corresponding z = NORMSINV(0.714) = 0.566
• S* = 600 + (0.566*52) = 629.43 units or 630 units

• Set a Level of Service (LOS) based on the Critical Ratio (CR) and set S* accordingly
• Whenever demand comes, i.e., order the same units demanded to fulfill the stock up to S* units
Continuous Review Policies

• Two-bin System / Order point, Order Quantity policy (s, Q)

• Policy: Order EOQ (Q*) when IP ≤ s (ROP)
• More popular

• Min-Max system / Order-Point, Order-Up-To-Level (s, S)

• Policy: Order (S-IP) if IP ≤ s

In both the cases, ROP (s) = μDL + z*σDL

z, in this case, is generally normsinv(Cycle Service Level)

• Example: You are managing the inventory for a production part with annual demand ~N(6,200, 800).
The cost of the item, c, is $100 and the holding charge is 15% per year. If ordering cost is $325 per
order, determine the economic order quantity (Q*). Lead time is 2 weeks. Assuming a CSL of 95%,
find the appropriate (s, Q) policy. Assume 52 weeks are there in a year.

Solution:

2∗𝐷∗𝑂𝐶 2∗6200∗325
Q* = = = 518.33 ≈ 518 units
ℎ∗𝑐 0.15∗100

Lead time is 2 weeks and so demand during lead time,

μDL = 6200*(2/52) = 6200/26 = 238.46 ≈ 238 units and,

σDL = 800* 2/52 = 157 units

Considering CSL = 0.95, z = 1.644 (see normal table or use excel sheet function to find z, z = normsinv(CSL)

s = 238 + (1.644*157) = 496.2 ≈ 496 units

So, the policy: Order 518 units when the stock is 496 units.
Periodic Review Policies
• Order up-to the level (R, S) • Hybrid (R, s, S) System
• Order (S – IP) at every R period • Policy: Every R time periods,
• Replenishment cycle system • Order S-IP if IP ≤ s,
• if IP>s then do not order
• General case for many policies
• Continuous Vs Periodic Review (A simple transformation)
(s, Q) (R, S)
s S
L R+L

• S = μ(DL + R) + z*σ(DL + R)
• z = normsinv(CSL)
• Order up to S at every R period
• Forecasted annual demand of hand drills is ~N(3,400, 400)
• Lead Time is 1 week
• Review Period is 4 weeks
• Desired CSL = 95%
• What would be the (R, S) policy? Assume 52 weeks are there in a year.

Solution:

μ(DL + R) = 3400*(4 + 1)/52 = 3400 / 10.4 = 326.9 ≈ 327 units

σ(DL + R) = 400*sqrt[(4+1)/52] = 400/ 10.4 = 124.03 ≈ 124 units

So, S = μ(DL + R) + z* σ(DL + R) = 327 + (1.64*124) = 530.4 ≈ 530 units

Policy: Order up to 530 units or (530 – IP) at every 4 weeks.

Freight Transportation
and Logistics

Arijit Mitra
Role of Transportation in Supply Chain
• Movement of product from one location to another

• Products rarely produced and consumed in the same location

• Significant cost component, but it is important because it gives the convenience and reach for customers

• Supply to the plant from different remote suppliers (Inbound Transportation)

• Sending the finished products to the downstream supply chain (Outbound Transportation)
• From plant to the warehouses or
• From warehouse to Wholesalers or
• From Wholesaler to retailers
• For centralization of inventories to fewer warehouses

• In short…to facilitate the movements of products from one point of Supply Chain to the other points of a
Supply Chain

• Shipper requires the movement of the product

• Carrier moves or transports the product

Modes of Transportation
• Air
• Package carriers
• Truck
• Rail
• Water
• Pipeline
• Intermodal
• Truck/Rail or Piggy-back
• Truck/Ship or Fishy-back
• Truck/Airways or Birdy-back
Transportation Facts: Performance of Different Modes

Freight Value Added to

Freight Value Freight Tons Ton-Miles GNP
($ billions) (millions) (billions) ($ billions)
Mode in 2011 in 2011 in 2011 in 2009
Air (includes truck and air) 394 6 11 61.9
Truck 12,181 11,924 2,337 113.1
Rail 588 2,053 1,518 30.8
Water 201 645 434 14.3
Pipeline 889 1,912 1,018 12.0
Multimodal 1,985 583 489
Air

• Cost components
1. Fixed infrastructure and equipment
2. Labor and fuel
3. Variable depending on passenger/cargo
• Key issues
• Location/number of hubs
• Fleet assignment
• Maintenance schedules
• Crew scheduling
• Prices and availability
Package Carriers

• Small packages up to about 150 pounds

• Expensive
• Rapid and reliable delivery
• Small and time-sensitive shipments
• Provide other value-added services
• Consolidation of shipments a key factor
Truck
• Significant fraction of the goods moved
• Truckload (TL)
• Low fixed cost
• Imbalance between flows
• Less than truckload (LTL)
• Small lots
• Hub and spoke system
• May take longer than TL
• Fatigue-related accidents
Rail

• Move commodities over large distances

• High fixed costs in equipment and facilities
• Scheduled to maximize utilization
• Transportation time can be long
• Trains ‘built’ not scheduled
Water

• Limited to certain geographic areas

• Ocean, inland waterway system, coastal waters
• Very large loads at very low cost
• Slowest
• Dominant in global trade
• Containers
Pipeline

• High fixed cost

• Primarily for crude petroleum, refined petroleum products, natural gas
• Best for large and stable flows
• Pricing structure encourages use for predicable component of demand
Intermodal

• Use of more than one mode of transportation to move a shipment

• Grown considerably with increased use of containers
• May be the only option for global trade
• More convenient for shippers – one entity
• Key issue – exchange of information to facilitate transfer between different
modes
• Piggy-back: It is a combination of rail and road. The containers are placed on railway flat-cars and
transported by rail from one terminal to another. After reaching the destination terminal, they can be
placed on trailers and transported by roads which are also called as TOFC that is Trailer on Flat Car or
COFC- Container on Flat Car

• Fishy-back: It is a combination of road and water transport. Fishy back/ train-ship/container-ship are
examples of the oldest mode of the intermodal transport. They utilize waterways, which are one of the
least expensive methods for line- haul movement. The fishy back, train ship and container ship concepts
load a truck trailer, railcar, or container on to ship for the line- haul move. Such services are provided in
coastal waters between Atlantic and Gulf ports, from the great lakes to coastal points and along inland
navigable waterways

• Birdy-back: It is a combination of road and airways and is generally used in International shipments.
Local cartage is a vital part of every air movement because air fright must eventually transport from the
airport to the final delivery destination. Air- truck movements usually provide service and flexibility
comparable to straight motor freight
Transportation Policies
• Either the government must build the rail and / or road infrastructure
and create a monopoly and necessary entry barriers so that private
companies can not enter this business
• If the government is not capable, it has to deregulate so that private
companies can compete each other to create these infrastructure
• Or a hybrid model like in some modes government’s presence and in
some other modes private company’s presence can work
Design Options for a Transportation Network

• When designing a transportation network

1. Should transportation be direct or through an intermediate site?
2. Should the intermediate site stock product or only serve as a cross-
docking location?
3. Should each delivery route supply a single destination or multiple
destinations?
Direct Shipment Network to Single Destination

• Transportation Problem

Figure 14-2 Direct Shipment Network

Direct Shipping with Milk Runs

Figure 14-3 Milk Runs from Multiple Suppliers or to Multiple Buyer Locations
All Shipments Via Intermediate Distribution Center with Storage

• Suppliers send their shipments to a

central distribution center
• Stored until needed by buyers
• Shipped to each buyer location
• Transshipment Problem
Storage • Usually known as Hub and Spoke
model

Figure 14-4 All Shipments via DC

All Shipments Via Intermediate Transit Point with Cross-Docking
• Suppliers send their shipments to an
intermediate transit point
• They are cross-docked and sent to buyer
locations without storing them

Only Cross
- docking
Shipping Via DC Using Milk Runs

Figure 14-5 Milk Runs from DC

 Pros and Cons of Different Transportation Networks

Network Structure Pros Cons

Direct shipping No intermediate warehouse High inventories (due
Simple to coordinate to large lot size)
Direct shipping with Lower transportation costs for small Increased coordination
milk runs lots Lower inventories complexity
All shipments via Lower inbound transportation cost Increased inventory
central DC with through consolidation cost Increased
inventory storage handling at DC
All shipments via Low inventory requirement Increased coordination
central DC with Lower transportation cost through complexity
cross-dock consolidation
Shipping via D C Lower outbound transportation cost Further increase in
using milk runs for small lots coordination
complexity
Tailored network Transportation choice best matches Highest coordination
needs of individual product and complexity
store
Tailored Network: Combinations of different Transportation Network

Depends on

• Customer Density
• Distance
• Size of the customer
• Product Demand
• Value of the product
Analysis of Freight
Transportation

Arijit Mitra
Selecting a transportation Network: A problem

• Eight stores to be supplied from four supply sources

• Truck capacity = 40,000 units
• Cost $1,000 per load, $100 per delivery
• Holding Cost - $0.2 per unit per year

The VP is thinking whether to use direct shipping from suppliers to retail stores or setting up milk runs
from suppliers to retail stores.

• What network should you recommend if annual sales for each product at each retail store are 960,000 units?

• What network should you recommend if annual sales for each product at each retail store are 120,000 units?
Thought before we start solving the problem

• First consideration – Should we investigate only transportation cost or

transportation and inventory cost?

• The concept of Total Least Lost! – Does it ring any bell?

• So, we need to calculate the total annual cost for each of the option

• Total Cost = Trucking Cost + Annual holding cost

• Trade-off between these two costs need to be analyzed

 Calculation for 960K Sales per year
𝑨𝒏𝒏𝒖𝒂𝒍 𝒔𝒂𝒍𝒆𝒔 = 𝟗𝟔𝟎, 𝟎𝟎𝟎/𝒔𝒕𝒐𝒓𝒆 𝑫𝒊𝒓𝒆𝒄𝒕 𝒔𝒉𝒊𝒑𝒑𝒊𝒏𝒈
Annual Trucking Cost (assuming 2 customers will be served at
Annual Trucking Cost: a time and each shipment will carry 20,000 units for each
customers i.e. A total of 40,000 units, a FTL)
(for the direct network, assumption is that each supplier delivers
40,000 unit to each of the customer because that is the FTL) Annual No. of shipments from each supplier to a particular
customer = Total Load / Truck Capacity
Annual No. of shipments from each supplier to a particular = 960,000 / 20,000 = 48
customer = Total Load / Truck Capacity
= 960,000 / 40,000 = 24 Annual Trucking Cost = Annual no. of shipments x
Transportation cost per shipment x no. of supplier x no. of
Annual Trucking Cost = Annual no. of shipments x Transportation customer
cost per shipment x no. of supplier x no. of customer = 48 x (1000/2 + 100) x 4 x 8 = $921,600
= 24 x (1000 + 100) x 4 x 8 = $844,800
Annual Holding cost = Average annual inventory / cycle
Annual Holding cost = Average annual inventory / cycle inventory inventory held at each supplier x holding cost rate x no. of
held at each supplier x holding cost rate x no. of suppliers suppliers
= (40,000/2) x 0.2 x 4 x 8 = $128,000 = (20,000/2) x 0.2 x 4 x 8 = $64,000

So, the Total cost for this option – ($844,800 + $128,000) = $972,800 So, the Total cost for this option – ($921,600 + $64,000) =
$985,600 (Little Higher!)
 Calculation for 120K Sales per year
𝑨𝒏𝒏𝒖𝒂𝒍 𝒔𝒂𝒍𝒆𝒔 = 𝟏𝟐𝟎, 𝟎𝟎𝟎/𝒔𝒕𝒐𝒓𝒆 𝑫𝒊𝒓𝒆𝒄𝒕 𝒔𝒉𝒊𝒑𝒑𝒊𝒏𝒈 𝑨𝒏𝒏𝒖𝒂𝒍 𝒔𝒂𝒍𝒆𝒔 = 𝟏𝟐𝟎, 𝟎𝟎𝟎/𝒔𝒕𝒐𝒓𝒆 𝑴𝒊𝒍𝒌 𝑹𝒖𝒏𝒔

Annual Trucking Cost (assuming 4 customers will be served at

Annual Trucking Cost: a time and each shipment will carry 10,000 units for each
customers i.e. A total of 40,000 units, an FTL)
(for the direct network, assumption is that each supplier delivers
40,000 unit to each of the customer because that is the FTL) Annual No. of shipments from each supplier to a particular
customer = Total Load / Truck Capacity
Annual No. of shipments from each supplier to a particular = 120,000 / 10,000 = 12
customer = Total Load / Truck Capacity
= 120,000 / 40,000 = 3 Annual Trucking Cost = Annual no. of shipments x
Transportation cost per shipment x no. of supplier x no. of
Annual Trucking Cost = Annual no. of shipments x Transportation customer
cost per shipment x no. of supplier x no. of customer = 12 x (1000/4 + 100) x 4 x 8 = $134,400
= 3 x (1000 + 100) x 4 x 8 = $105,600
Annual Holding cost = Average annual inventory / cycle
Annual Holding cost = Average annual inventory / cycle inventory inventory held at each supplier x holding cost rate x no. of
held at each supplier x holding cost rate x no. of suppliers suppliers
= (40,000/2) x 0.2 x 4 x 8 = $128,000 = (10,000/2) x 0.2 x 4 x 8 = $32,000
So, the Total cost for this option – ($105,600 + $128,000) = $233,600 So, the Total cost for this option – ($134,400 + $32,000) =
$166,400 (Quite lower this time!)
 Calculating Total cost including inventory and Transportation

•The earlier case was only considering the cycle stock cost

•In practice, we should also consider another two components namely, safety stock and pipeline
inventory

•Total cost = Purchasing Cost + Ordering cost + Holding cost of all the three components of
stock + transportation cost

•TC = c*D + (D/Q*) x OC + h x (Q*/2 + z*SS + D*L) + Transportation Cost

•We call it total logistics cost!

Trade-off in Transportation: Trade-off among the transportation modes

• Here, we calculate the total logistics cost (we may ignore the purchase and ordering cost because these
two components are same for all the modes and so, they are not relevant costs!)

• See the problem of Eastern Electric (EE) in the book

• Demand = 120,000 motors

• Cost = $120/motor
Transportation Proposals for EE Electric
• Weight = 10 lbs/motor
• Lot size = 3,000
Range of Quantity
• Safety stock = 50% ddlt
Carrier Shipped (cwt) Shipping Cost ($/cwt)
• Golden is offering marginal AM Railroad 200+ 6.50
discount with more weight Northeast Trucking 100+ 7.50
• 1 cwt = 100 lbs Golden Freightways 50–150 8.00
• Calculate the total logistics cost
Golden Freightways 150–250 6.00
for each options and select the
Golden Freightways 250+ 4.00
cheapest option!
Other Trade-offs in transportation

• Aggregation of inventory
• Aggregate the inventories required in all the territories and keep it in a central warehouse
(ultimately a trade-off between holding cost and transportation)

• Transportation cost generally increases but the holding cost decreases due to reduction in
the safety inventory

• Trade-off between transportation cost & responsiveness

• Date of Delivery

• Level of Service
Tradeoffs When Aggregating Inventory: HighMed Problem
(See in book)

HighVal – weekly demand H = 2, H = 5, weight = 0.1 lbs, cost = $250

LowVal – weekly demand = 20, = 5, weight = 0.04 lbs, cost = $30
L L

CSL = 0.997, holding cost = 25%, L = 1 week, T = 4 weeks

UPS lead time = 1 week, $0.66 + 0.26x
FedEx lead time = overnight, $5.53 + 0.53x
Option A: Keep the current structure but replenish inventory
once a week rather than once every four weeks
Option B: Eliminate inventories in the territories, aggregate
all inventories in a finished-goods warehouse at Madison,
and replenish the warehouse once a week