MCL756 : Supply Chain Management

3-0-0 Course

Dr S G Deshmukh & Prof Prem Vrat

1
 Jan22T Tue Wed Fri

Today is 12th Feb 2025 Week 1 3 Jan(SGD)

Week 2 7 8 10
Week 3 16 15 17
Week 4 21 22 24(SGD)
Week 5 28(PV) 29 31
Week 6 4 Feb 7
Week 11 Feb 12 Feb
7(feb)
Week 8
Week 9
Week 10
Week 11

Week 12
Week 13
Week 14

What was covered in previous Transportation driver (contd)

Week 15
Week 16

session? Week17

What will be done in this session? Distribution strategies

 Cross docking- Indian cases

◦ TCI has its trucks that aggregate LTL shipments requiring delivery at
other places from smaller towns like Hubli, Mangalore and Bellary.
 All the trucks from the smaller towns of Karnataka will arrive at
Bangalore hub and the good from these trucks will be cross-
docked into four trucks leaving for Mumbai/Chennai/Delhi and
Kolkata
 Maruti, Toyota and Tata Motors have started using concept of
cross docking.
 Totyota- Cross docking is done at Gurugram and Pune

3 3
 Direct Shipping

4
 Comparison of Distribution Network
Design Options: Illustration
 Manufacturing firm has three plants (A, B &C), each manufacturing a different
product line and serving a stable market through three depots ( X,Y &Z).
 Plant A is manufacturing menswear, plant B is manufacturing ladies wear and plant
C is manufacturing children’s wear.
 Weekly demand = 100 units for each of the three types of garments at each of
the three depots
 Truck can carry 300 units of garments and the transport cost is Rs 2 per km. for
TL shipments. To obtain economies of scale firm has decided to work with TL
shipments.. Inventory carrying cost is at 20% per annum.
 All the products cost Rs 200 per unit, so inventory carrying cost is Rs 40 per unit
per year.
 Facility cost of maintaining a DC is Rs 12,000 per year.

5
 Transportation Strategy:
Linking Plants to Markets

6
 Comparison of the Three
Transportation Strategies

7
 Cost computation
 Distance travelled per cycle =
2XA+2XB+2XC+2YA+2YB+2YC+2ZA+2ZB+2ZC= 3744 Km
 Travel cost per cycle = 3744 x2 =7488
 Number of cycles per year = 52/3
 Annual transport cost = 7488 x 52/3 = 129,781
 In direct shipment, each depot receives a shipment of 300 units
each products, so average inv = 150 units. This results in 450
garments of cycle stock at each depot or 3 x 450= 1350 units
across 3 depots
 Thus the average inv cost = 1350 x 40 = Rs 54,000

8
 Tailored Transportation
 The use of different transportation networks and modes based
on customer and product characteristics
 Factors affecting tailoring:
◦ Customer distance and density
◦ Customer size
◦ Product demand and value

9
 Routing and Scheduling
in Transportation
 The most important operational decision related to
transportation in a supply chain is the routing and scheduling of
deliveries
 Decision of which customers to be visited by a particular vehicle
and the sequence in which they will be visited
 basic approaches:
◦ Savings matrix method

10
 Vehicle Scheduling Problem
 The manager then repeats the process with three vehicles, one to
the South side, one to the East side, and one to the West side. This
plan appeared suitable to meet all the due dates.
 Such problems may appear in case of e.g., an on-line grocery stores
such as http://www.peapod.com.
 The orders are taken 24 hours on-line. When the manager starts
working in the morning, the manager has to decide number of
vehicles and their routes.

11 11
 12

12
 Vehicle Scheduling Problem
 The objectives of the vehicle scheduling problem can be many.
 Following are some examples:
◦ Minimize number of vehicles required
◦ Minimize total distance traveled
◦ Minimize total travel time
◦ Minimize cost
 We shall consider the problem with the objective of minimizing total
distance.

13 13
 Vehicle Scheduling Problem

X Coordinate Y Coordinate Order Size

W 0 0
1 0 12 48
2 6 5 60
3 7 15 43
4 9 12 92
5 15 3 80

14 14
 Vehicle routing
 The Vehicle Routing Problem is the problem of constructing
a set of tours from a depot to demand points such that the
sum of the lengths is minimized.
 Usually, there are capacity constraints on the tours, or the
number of tours is prescribed.

 Solution approaches: …
◦ Savings algorithm

15
 Vehicle Routing Problem

16 16/58
 Vehicle Routing Problem

17
 The savings algorithm
 The savings algorithm is a simple and very fast algorithm.
 It assumes that each route has a certain capacity, and each
demand point has a certain volume.
 First, compute the savings from each connection i to j,
s_ij = c_i0 + c_0j – c_ij.
 Order the edges on their savings from big to small.
 Add edges to routes until no more can be added.
This is also called as Clarke and Wright Algorithm
 G. Clarke, J. Wright, “Scheduling of vehicles from a central depot to a number of delivery points”, Operations Research, Vol. 12, No. 4, pp. 568-581,
1964 DOI: https://doi.org/10.1287/opre.12.4.568

18
 Example- A case of Public Logistics
Company
 Distribution of rice

 Tiara Risa Damayanti, Ade Lita Kusumaningrum, Yulia Dwi Susanty, and Sri Susilawati
Islam , Route Optimization Using Saving Matrix Method – A Case Study at Public Logistics Company in Indonesia,
Proceedings of the 5th NA International Conference on Industrial Engineering and Operations Management ,Detroit,
Michigan, USA, August 10 - 14, 2020

19
20
21
New Routes..
Number reduced to 9 routes from original
18
 Expected savings
Proposed Old Route Savings Savings in %
Distance 1638.4 Km 886.2 Km 752.2 Km 45.91%
travelled
(Km)
Cost 1925,120 1041,285 888,835 45.91%

23
 Vehicle Routing Problem
 For each customer, we know
 Quantity required
 The cost to travel to every other customer

 For the vehicle fleet, we know

 The number of vehicles
 The capacity

 We must determine which customers each vehicle serves,

 and in what order, to minimise cost

24
 Vehicle Routing Problem

 Objective function
 In academic studies, usually a combination:
◦ First, minimise number of routes
◦ Then minimise total distance or total time

 In real world
◦ A combination of time and distance
◦ Must include vehicle- and staff-dependent costs
◦ Usually vehicle numbers are fixed
25
 Vehicle Routing Problem
• Might be more than one ‘capacity’
– e.g. limited weight and volume
• Might be more than one type of vehicle
– Different capacities and costs
– ‘Heterogeneous Vehicles’

26
 Savings Matrix Method
 Identify the distance matrix
 Identify the savings matrix
 Assign customers to vehicles or routes
 Sequence customers within routes

27
 Savings Method of Routing
Initial routing – Route distance
= d0,A + dA,0 + d0,B + dB,0 d0,A

Stop A
dA,0

Depot (0)
dB,0
Stop B

d0,B

28
 Savings Method of Routing
Combing 2 stops on 1 route – Route distance
= d0,A + dA,B + dB,0

d0,A
Stop A

dA,B
Depot (0)

dB,0
Stop B

Savings value of S
= d0,A + dB,0 - dA,B
29
 Savings Method of Routing
Initial routing – Route distance
= d0,A + dA,0 + d0,B + dB,0 d0,A = 100

A (A)
dA,0 = 100

J (0)
dB,0 = 85
B (B)

d0,B = 85

Initial routing – Route distance

30 = d0,A(100) + dA,0(100) + d0,B(85) + dB,0(85) = 370 Kms total
 Savings Method of Routing
Combining 2 stops on 1 route – Route distance
= d0,A(100) + dA,B(145) + dB,0(85) = 330 total Kms

d0,A = 100 A(A)

dA,B = 145
J (0)

dB,0 = 85
B (B)

Savings value of S
= d0,A(100) + dB,0(85) - dA,B(145) = 40 Kms saved
31
 Vehicle Scheduling Problem
 The manager must solve two sub-problems
1. Split the city into several smaller regions, each of which will
be served by one vehicle. This can be done by considering a
customer first, assigning the customer to a vehicle, and then
assigning other nearby customers to the same vehicle. So,
this sub-problem will be called assigning customers to
vehicles.
2. Sequence customers served by the same vehicle.

32
 Vehicle Routing Problem
 The objectives of the vehicle scheduling problem can be
many.
 Following are some examples:
◦ Minimize number of vehicles required
◦ Minimize total distance traveled
◦ Minimize total travel time
◦ Minimize cost
 We shall consider the problem with the objective of
minimizing total distance.

33
 Assume that Example
There are orders from 5 different customers
There are 2 trucks each capable of carrying 200 units

X Coordinate Y Coordinate Order Size

W 0 0
1 0 12 48
2 6 5 60
3 7 15 43
4 9 12 92
5 15 3 80

Assume that the customer locations and order sizes

are as shown above. The locations are plotted on the
34 next slide.
 Identify Distance Matrix
• First, the Euclidean distances are computed. The formula
and a sample computation is shown below. The other
distances are computed similarly and shown on the next
slide.

Dist( A, B) = (x A − xB )2 + ( y A − yB )2

Example :
Dist(1,2) =

35 35
 Distance Matrix

Distance Matrix
W Cust 1 Cust 2 Cust 3 Cust 4 Cust 5
Warehouse 0 12.0 7.8 16.6 15.0 15.3
Customer 1 0 9.2 7.6 9.0 17.5
Customer 2 0 10.0 7.6 9.2
Customer 3 0 3.6 14.4
Customer 4 0 10.8
Customer 5 0

36
 Identify Savings Matrix
• Instead of serving two different customers by two different vehicles, if
a single vehicle is used to serve both the customers, then some
traveling distance is saved.
• Savings S(A,B) represents the savings in traveling distance obtained
by assigning Customers A and B to the same vehicle instead of
assigning them to two different vehicles.
• The capacity constraint is not considered at the time of computation of
savings. In other words, when computing savings, assume an infinite
capacity of the vehicles.

37
 Identify Savings Matrix

Savings Matrix
Cust 1 Cust 2 Cust 3 Cust 4 Cust 5
Customer 1 0 10.6 20.9 18.0 9.8
Customer 2 0 14.3 15.2 13.9
Customer 3 0 27.9 17.4
Customer 4 0 19.5
Customer 5 0

38
 Rank Savings
• The next step is to rank the savings. The idea is to merge those two
customers to the same vehicle, whose merging gives the highest
savings.
• The savings are ranked from high to low.
• From the savings matrix shown on the previous slide, the highest
savings of 27.9 is obtained by merging Customers 3 and 4 to the same
vehicle.
• Next highest savings of 20.9 is obtained by merging Customers 1 and 3
to the same vehicle.
• Similarly the other savings are ranked and shown on the next slide.

39
 Assign Customers to Vehicles
Next, merge the
customers. The
pair giving the 3 Order
Customer
Size
highest savings is
1 4 1 48
merged first if the
2 60
capacity is
3 43
available.
5 4 92
2 5 80
W
Location of Warehouse and Customers

Rank (3,4) (1,3) (4,5) (1,4) (3,5)

40 (2,4) (2,3) (2,5) (1,2) (1,5)
 Assign Customers to Vehicles

To merge the
lowest ranked
pair (3,4), the 3 Order
Customer
Size
capacity
1 4 1 48
required =
2 60
43+92= 135 <
3 43
200 = capacity
5 4 92
available. So, 2 5 80
merge 3 and 4. W

Location of Warehouse and Customers

Rank (3,4) (1,3) (4,5) (1,4) (3,5)

41 (2,4) (2,3) (2,5) (1,2) (1,5)
 Assign Customers to Vehicles
To merge the next pair
(1,3), capacity required
= 43+92+40= 175 < 200 3 Order
Customer
Size
= capacity available. So,
1 4 1 48
merge 1 and 3 (and 4).
2 60
3 43
5 4 92
2 5 80
W
Location of Warehouse and Customers

Rank (3,4) (1,3) (4,5) (1,4) (3,5)

42 (2,4) (2,3) (2,5) (1,2) (1,5)
 Assign Customers to Vehicles
Merging (4,5), (3,5), (2,4)
and (2,3) requires more
capacity than available. 3 Order
Customer
Size
The pair (1,4) is already
1 4 1 48
merged. So, the pairs are
2 60
crossed out.
3 43
5 4 92
2 5 80
W
Location of Warehouse and Customers

Rank (3,4) (1,3) (4,5) (1,4) (3,5)

43 (2,4) (2,3) (2,5) (1,2) (1,5)
 Assign Customers to Vehicles
The next pair (2,5) are
merged and assigned to a
new vehicle as the capacity 3 Order
Customer
available = 200 > 60 + 80 = Size
140 = capacity required. 1 4 1 48
2 60
3 43
5 4 92
2 5 80
W
Location of Warehouse and Customers

Rank (3,4) (1,3) (4,5) (1,4) (3,5)

44 (2,4) (2,3) (2,5) (1,2) (1,5)
 Sequence Customers
 The next step is sequencing customers assigned to the same vehicle. A
question is in what sequence will the first vehicle visit customers 1, 3 and
4 and return to the warehouse? Similarly, another question is in what
sequence will the other vehicle visit customers 2 and 5.
 This problem is popularly called the traveling salesman problem.
 We shall use the nearest neighbor rule which states that always visit the
customer that is nearest.

45
 Sequence Customers
First, consider the problem of 3
sequencing customers 1, 3 and
1 4
4 who are assigned to the
same vehicle. The relevant
distances are copied from the 5
distance matrix and shown 2
below. W
Location
Distance Matrix
W Cust 1 Cust 3 Cust 4
Warehouse 0 12.0 16.6 15.0
Customer 1 0 7.6 9.0
Customer 3 0 3.6
46 Customer 4 0
 Sequence Customers
Among 1, 3 and 4 is 1 is the nearest to
Warehouse. So, the vehicle will first 3
travel from Warehouse to customer 1. 4
1
The row (=from) corresponding to
Warehouse and the column (=to)
corresponding to customer 1 are 5
crossed out. 2
W
Location
Distance Matrix
W Cust 1 Cust 3 Cust 4
Warehouse 0 12.0 16.6 15.0
Customer 1 0 7.6 9.0
Customer 3 0 3.6
47 Customer 4 0
 Sequence Customers
Between 3 and 4 is 3 is the 3
nearest to 1. So, the vehicle will 4
1
travel from 1 to 3. The row (=from)
corresponding to customer 1 and
the column (=to) corresponding to 5
customer 3 are crossed out. 2
W
Location
Distance Matrix
W Cust 1 Cust 3 Cust 4
Warehouse 0 12.0 16.6 15.0
Customer 1 0 7.6 9.0
Customer 3 0 3.6
48 Customer 4 0
 Sequence Customers
The only possible tour is then W-1- 3
3-4-W. Next, consider the problem 4
1
of sequencing customers 2 and 5
who are assigned to the same
vehicle. The relevant distances are 5
copied from the distance matrix 2
and shown below. W
Location
Distance Matrix
W Cust 2 Cust 5
Warehouse 0 7.8 15.3
Customer 2 0 9.2
Customer 5 0
49
 Sequence Customers
Between 2 and 5 is 2 is the nearest to the 3
Warehouse. So, the vehicle will travel from
1 4
the Warehouse to customer 2. The only tour
is then W-2-5-W. Note: due to symmetry both
W-2-5-W and W-5-2-W have the same 5
distance traveled. 2
W
Location
Distance Matrix
W Cust 2 Cust 5
Warehouse 0 7.8 15.3
Customer 2 0 9.2
Customer 5 0
50
 Traveling Salesperson Problem

51
52
 TSP Construction Heuristics
 Nearest neighbor.
◦ Add nearest customer to end of the route.

 Savings method.
◦ Add customer that saves the most to the route.

53
 Nearest Neighbor
E
A B  Add nearest customer to end of the route.
C
D

depot E
F E
E A B
A B A B
Depot A B C D E F
C C D C
Depot --- 2 5 8 6 9 7 D
D
A -- 3 7 5 8 6 F F
B -- 1 3 4 5 F
depot
depot depot
C -- 7 9 8
D -- 5 9 1 2 3
E -- 7
54 F
 Nearest Neighbor
E
A B  Add nearest customer to end of the route.
C
D

depot E
F E
E A B
A B A B
Depot A B C D E F
C C D C
Depot --- 2 5 8 6 9 7 D
D
A -- 3 7 5 8 6 F F
B -- 1 3 4 5 F
depot depot depot
C -- 7 9 8
D -- 5 9 4 5 6
E -- 7
55 F
 Savings Method
1. Select any city as the “depot” and call it city “0”.
- Start with separate one stop routes from depot to each customer.

2. Calculate all savings for joining two customers and eliminating a trip back to
the depot.
Sij = Ci0 + C0j - Cij

3. Order savings from largest to smallest.

4. Form route by linking customers according to savings.

- Do not break any links formed earlier.
- Stop when all customers are on the route.

56
 Savings Method Example

Given 5 customers and the costs (distances) between them.

2 3
1 Cij 0 1 2 3 4
0 -- 8 9 13 10
1 8 -- 4 11 13
4 2 9 4 -- 5 8
3 13 11 5 -- 7
0 4 10 13 8 7 --

57
 Savings Method Example

Given 5 customers, select Conceptually form routes from

the lower left as the depot. the depot to each customer.

2 3
1 2 3
1

4
4

0 depot

58
 Savings Method: S12
Remove Add
Savings = S12 = C10+C02 - C12
3 = 8 + 9 - 4 = 13
1 2 3
1 2

4
4

depot
Cij 0 1 2 3 4
depot
0 -- 8 9 13 10
1 8 -- 4 11 13
2 9 4 -- 5 8
3 13 11 5 -- 7
4 10 13 8 7 --
59
 Savings Method

S12 = C10 + C02 - C12 If problem is symmetric, then

sij = sji, so s21 = s12, s32 = s23,
Note: S21 = C20 + C01 - C21 etc. There are (n-1)(n-2)/2
so S12 = S21 savings to calculate.

2 3
1

If problem is asymmetric,
then all sij‘s must be calculated.
4 There are (n-1)(n-2) savings
to calculate.
depot

60
 Savings Method: S13
S13 = C10+C03 -C13
= 8 + 13 - 11 = 10
3
Cij 0 1 2 3 4
1 2
0 -- 8 9 13 10
1 8 -- 4 11 13
2 9 4 -- 5 8
4
3 13 11 5 -- 7

depot 4 10 13 8 7 --

61
 Savings Method: S14
S14 = C10+C04 -C14
= 8 + 10 - 13 = 5

2 3
1

Cij 0 1 2 3 4

4
0 -- 8 9 13 10
1 8 -- 4 11 13
depot 2 9 4 -- 5 8
3 13 11 5 -- 7
4 10 13 8 7 --

62
 Savings Method: S23
S23 = C20+C03 -C23
= 9 + 13 - 5 = 17

2 3
1

Cij 0 1 2 3 4

4
0 -- 8 9 13 10
1 8 -- 4 11 13
depot 2 9 4 -- 5 8
3 13 11 5 -- 7
4 1- 13 8 7 --

63
 Savings Method: S24
S24 = C20+C04 -C24
= 9 + 10 - 8 = 11
3
Cij 0 1 2 3 4
1 2
0 -- 8 9 13 10
1 8 -- 4 11 13
2 9 4 -- 5 8
4
3 13 11 5 -- 7

depot 4 10 13 8 7 --

64
 Savings Method: S34
S14 = C30+C04 -C34
= 13 + 10 - 7 = 16

2 3
1 Cij 0 1 2 3 4
0 -- 8 9 13 10
1 8 -- 4 11 13
4 2 9 4 -- 5 8
3 13 11 5 -- 7
depot 4 10 13 8 7 --

65
 Savings Method

 Order savings from largest to smallest.

S23 (= S23) = 17
S34 (= S43) = 16
S12 (= S21) = 13
S24 (= S42) = 11
S13 (= S31) = 10
S14 (= S41) = 5

66
 Savings Method

 Form route by linking customers

according to savings.
Link 2 and 3.
S23
S34
2 3
S12 1

S24
S13 4

S14 depot
67
 Savings Method

 Form route by linking customers

according to savings.

S23 0-2-3-0
S34 3
1 2
S12
S24
S13 4

S14 depot

68
 Savings Method

 Form route by linking customers

according to savings.

Link 3 and 4.
S23 0-2-3-0 Do not break earlier links.
S34 3
1 2
S12
S24
S13 4

S14 depot

69
 Savings Method

 Form route by linking customers

according to savings.

S23 0-2-3-0
S34 0-2-3-4-0
S12 1 2 3

S24
S13 4

S14
depot
70
 Savings Method

 Form route by linking customers

according to savings.

S23 0-2-3-0 Link 1 and 2.

Do not break earlier links.
S34 0-2-3-4-0
S12 1 2 3

S24
S13 4

S14
depot
71
 Savings Method

 Form route by linking customers

according to savings.

S23 0-2-3-0 Done!

S34 0-2-3-4-0
S12 0-1-2-3-4-01 2 3

S24
S13 4

S14
depot
72
 VRP in real life ..1..
 Many groups now looking at real-world constraints
 Rich Vehicle Routing Problem
• Attempt to model constraints common to many real-life
enterprises
– Multiple Time windows
– Multiple Commodities
– Heterogeneous vehicles
– Compatibility constraints
• Goods for customer A can’t travel with goods from customer B
• Goods for customer A can’t travel on vehicle C
73
 VRP in real life ..2..
 Other real-world considerations
• Fatigue rules and driver breaks
• Vehicle re-use
• Ability to change vehicle characteristics (composition)
– Add trailer, or move compartment divider
• Use of limited resources
– e.g limited docks for loading, hence need to stagger dispatch times
• Variable loading /
unloading times
74 74/58
 VRP in real life ..3..
 Yet more constraints
• Only two types of product on each vehicle
• ‘Grand tour’ constraint
– Customers visited in ‘patterns’ (Period Routing)
– Grand tour includes all visits
– Skip visit if not required that day
– Each day must be feasible (incl. time windows)
• Meet ferry
• Blood transport (dynamic time window)
• Driver constraint
75
 VRP in real life ..4..
 New data sources

• Routing with time-of-day dependent travel times

– Uses historical data to forecast travel time at different times of day

• Routing with dynamic travel times

– Uses live traffic information feed to update expected travel time
dynamically

76
 VRP in real life ..5..
 Stochastic Routing
 What if things don’t go according to plan?
• Sources of uncertainty
– Uncertainty in existence (do I even need to visit)
– Uncertainty in quantity (how much is actually required)
– Uncertainty in travel times (traffic)
– Uncertainty in duration (maintenance engineer)

• Optimal solution can be brittle

– If something is not quite right, whole solution falls apart

77
 Summary-I
 There are strong economies of distance and economies of scale in
the transport industry.
 Value density and demand characteristics of the product strongly
influence transportation decisions. is vital for supply chain mangers.
Poor understanding of transportation cost drivers can have serious
business implications for firms.
 Firms insisting on FTL shipments may get transportation
efficiencies but might end with overall increase in cost because of
increased level of inventories. Transportation decisions should not
be looked at in isolation

78
 Summary-II
 Innovative ideas like milk run and cross docking help firms in managing
frequent delivery from suppliers without sacrificing transportation
efficiencies.
 Transportation cost has become quite a serous issue in the last few
years because fuel prices have shot up considerably and LTL supplies can
add significant costs to firms.
 Progressive firms like Wal-Mart and Tesco have paid a lot of attention to
their transportation strategy and use a mix and match of all the
transportation strategies discussed in this section.
 Transportation decisions play an extremely important role for e-
retailers. Because of the high costs involved in last mile transportation to
the customer’s home, retailing on web is likely to be extremely beneficial
only for firms which have high product variety, high value density and
high demand uncertainty.
79