ISM-C1004 - Business Analytics 1 - Assignment 2 (Total 36 points)

Transportation problem (4pts)

CaaS Ltd (Car-as-a-service) offers electric cars for rental in Helsinki. A customer
can take a car into use with a phone app, and the customer is charged based
on the time the car is in use. Customers can also leave the car anywhere in
Helsinki.

CaaS has divided Helsinki into six areas (A-F), for each of which it has good
predictions on the demand, i.e., how many cars are needed in that area each
day (Table 1). CaaS wants to build a decision support model that identifies the
optimal way to redistribute the cars during the night so that each area has
enough cars to satisfy the demand the next day. CaaS has also estimated the
average travel times between each of these areas (Table 2). The current
locations of the cars are given in Table 3.

Implement on the spreadsheet a network LP model that minimizes the total

time it takes to redistribute the cars so that demand is satisfied.
Find the optimal solution. Report the number of cars transferred between Grading
each area, i.e., prepare a list of transfers.
Model is correct, correct opti
List of transfers: presented (decision varibles
hidden in formulas but prese
From A to E = 5 function and constraint coeffi
From B to C = 5 Small error in the model or i
From B to F = 3 Model not clearly presented
From D to F = 7 There are major errors in the
Model is poorly presented (-
Report also the total driving time that the optimal reallocation requires.
No spreadsheet implementa
Not a linear model (0 pts)
Total driving time: 423
Table 1. Demand
Area A B C D E F

Demand 20 27 20 28 25 20

Table 2. Travel time in minutes

To area
From area A B C D E F
A - 20 28 29 17 32
B 23 - 17 31 28 26
C 23 17 - 20 29 16
D 14 26 23 - 20 25
E 23 34 26 14 - 17
F 38 20 16 28 25 -

Table 3. Current location of the fleet

Area A B C D E F
Cars 25 35 15 35 20 10
(Total 36 points)
Spreadsheet implementation her

from area

objective function

Grading constraints
Model is correct, correct optimal solution is reported and the model is clearly
presented (decision varibles, data, and formulas clearly marked; data not
hidden in formulas but presented in data cells corresponding to objective
function and constraint coefficients). (4 pts)
Small error in the model or in reporting solutions (-1 pt)
Model not clearly presented (-1 pt)
There are major errors in the model (-2 to -3 pts)
Model is poorly presented (-2 pts)

No spreadsheet implementation given (0 pts)

Not a linear model (0 pts)
eet implementation here decision variables: number of cars transported to area
a b c d e f
a 20 0 0 0 5 0
b 0 27 5 0 0 3
c 0 0 15 0 0 0
d 0 0 0 28 0 7
e 0 0 0 0 20 0
f 0 0 0 0 0 10
0 20 28 29 17 32 423
23 0 17 31 28 26
23 17 0 20 29 16
min
14 26 23 0 20 25
23 34 26 14 0 17
38 20 16 28 25 0
demand
1 1 1 1 1 1 25 <=
1 1 1 1 1 1 35 <=
1 1 1 1 1 1 15 <=
capacity
1 1 1 1 1 1 35 <=
1 1 1 1 1 1 20 <=
1 1 1 1 1 1 10 <=
20 27 20 28 25 20
= = = = = =
20 27 20 28 25 20
25
35
15
35
20
10
ISM-C1004 - Business Analytics 1 - Assignment 2 (Total 36 points)
Assignment problem (5pts) Grading
FIXITNOW is a small company that repairs consumer electronics such as phones,
tablets and computers. Model is correct, c
presented (decisio
The currently ordered jobs and the time required to complete them are listed in hidden in formula
Table 1. The acronyms for the job types stand for Windows, Android, IOS, Phone, function and cons
Tablet and Computer. For instance, CW denotes a Computer running Windows. Small error in the
Model not clearly
The engineers of FIXITNOW, their salaries per hour and their skills are listed in There are major e
Table 2. Each engineer works at most 12 hours and can only work on jobs for Model is poorly pr
which (s)he has the necessary skills. No spreadsheet im
Not a linear mode
Implement on the spreadsheet a LP-model that assigns engineers' work hours to
the jobs so that total costs are minimized and all jobs are done. [HINT: you can
set a very high cost in the objective function to assignments that are not feasible]

Report the engineers that do not get the maximum number of hours in the
optimal solution.

Engineers that do not get the maximum number of hours: Leo,

Väinö, Aada
Report the optimal total cost.

Optimal total cost: 2987

Table 2 Table 1
Engineer Salary Skills Job Type Hours
Eeva 32 TA, TI, CW #1 CI 6
Emma 32 PI, TI, CW #2 CW 8
Leo 34 TA, TW #3 PA 2
Väinö 34 TI #6 PI 1

Onni 32 CI, CW #7 TI 5
Aada 31 PA #8 PA 2
Ellen 29 CW, CI #9 TA 5
#10 CI 4
#11 PI 2
#12 TI 5
#13 CW 5
#14 PA 5
#15 TA 4
#16 CI 4
#17 PI 3
#18 CW 7
#19 TA 3
#20 CW 4
#21 PI 7
#22 TA 6
Total 36 points)
Grading
Model is correct, correct optimal solution is reported and the model is clearly
presented (decision varibles, data, and formulas clearly marked; data not
hidden in formulas but presented in data cells corresponding to objective
function and constraint coefficients). (5pts)
Small error in the model or in reporting solutions (-1 pt) Decision variables: Working h
Model not clearly presented (-1 pt)
There are major errors in the model (-2 to -3 pts)
Model is poorly presented (-2 pts) Eeva
No spreadsheet implementation given (0 pts) Emma
Not a linear model (0 pts) Leo
Elias
Sofia
Väinö
Onni
Aada
Ellen
Feasible job
worker
Eeva
Emma
Leo
Elias
Sofia
Väinö
Onni
Aada
Ellen
Coefficient
worker
Eeva
Emma
Leo
Elias
Sofia
Väinö
Onni
Aada
Ellen

contraints

Eeva
Emma
Leo
Elias
working hour Sofia
for each job
Väinö
working hour
for each job
Onni
Aada
Ellen

working hour
 Spreadsheet implementation here
Decision variables: Working hours of each employee on each job
#1 #2 #3 #4 #5 #6 #7 #8 #9 #10
0 0 0 1 0 0 0 0 5 0
0 1 0 0 0 0 1 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 1 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 4 0 0 0
1 7 0 0 0 0 0 0 0 4
0 0 2 0 0 0 0 2 0 0
5 0 0 0 3 0 0 0 0 0
Feasible job
#1 #2 #3 #4 #5 #6 #7 #8 #9 #10
n y n y n n y n y n
n y n n n y y n n n
n n n y n n n n y n
n n n n n y y n n n
n y n n n n n n n n
n n n n n n y n n n
y y n n y n n n n y
n n y n n n n y n n
y y n n y n n n n y
Coefficient
#1 #2 #3 #4 #5 #6 #7 #8 #9 #10
1000 32 1000 32 1000 1000 32 1000 32 1000
1000 32 1000 1000 1000 32 32 1000 1000 1000
1000 1000 1000 34 1000 1000 1000 1000 34 1000
1000 1000 1000 1000 1000 31 31 1000 1000 1000
1000 31 1000 1000 1000 1000 1000 1000 1000 1000
1000 1000 1000 1000 1000 1000 34 1000 1000 1000
32 32 1000 1000 32 1000 1000 1000 1000 32
1000 1000 31 1000 1000 1000 1000 31 1000 1000
29 29 1000 1000 29 1000 1000 1000 1000 29

working hour for

#1 #2 #3 #4 #5 #6 #7 #8 #9 #10
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
6 8 2 1 3 1 5 2 5 4
= = = = = = = = = =
6 8 2 1 3 1 5 2 5 4
 #11 #12 #13 #14 #15 #16 #17 #18 #19 #20
0 0 0 0 0 0 0 0 3 0
0 0 0 0 0 0 3 4 0 0
0 0 0 0 4 0 0 0 0 0
2 5 0 0 0 0 0 0 0 0
0 0 5 0 0 0 0 3 0 4
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 5 0 0 0 0 0 0
0 0 0 0 0 4 0 0 0 0

#11 #12 #13 #14 #15 #16 #17 #18 #19 #20
n y y n y n n y y y
y y y n n n y y n y
n n n n y n n n y n
y y n n n n y n n n
n n y n n n n y n y
n y n n n n n n n n
n n y n n y n y n y
n n n y n n n n n n
n n y n n y n y n y

#11 #12 #13 #14 #15 #16 #17 #18 #19 #20
1000 32 32 1000 32 1000 1000 32 32 32
32 32 32 1000 1000 1000 32 32 1000 32
1000 1000 1000 1000 34 1000 1000 1000 34 1000
31 31 1000 1000 1000 1000 31 1000 1000 1000
1000 1000 31 1000 1000 1000 1000 31 1000 31
1000 34 1000 1000 1000 1000 1000 1000 1000 1000
1000 1000 32 1000 1000 32 1000 32 1000 32
1000 1000 1000 31 1000 1000 1000 1000 1000 1000
1000 1000 29 1000 1000 29 1000 29 1000 29

working hour for each employee

#11 #12 #13 #14 #15 #16 #17 #18 #19 #20
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
2 5 5 5 4 4 3 7 3 4
= = = = = = = = = =
2 5 5 5 4 4 3 7 3 4
 #21 #22
0 3
3 0
0 3
4 0
0 0
0 0
0 0
0 0
0 0

#21 #22
n y
y n
n y
y n
n n
n n
n n
n n
n n

#21 #22
1000 32
32 1000
1000 34
31 1000
1000 1000
1000 1000
1000 1000
1000 1000
1000 1000
min 2897

#21 #22 LHS RHS

1 1 12 <= 12
1 1 12 <= 12
1 1 7 <= 12
1 1 12 <= 12
1 1 12 <= 12
1 1 4 <= 12
 1 1 12 <= 12
1 1 9 <= 12
1 1 12 <= 12
7 6
= =
7 6
ISM-C1004 - Business Analytics 1 - Assignment 2 (Total 36 points)
Transshipment model (7 pts)
Pulp & Paper Ltd. produces carton at 3 mills in Finland, Sweden and Belgium. The monthly production capacities are give
the cartboard is transported to 4 warehouses located in Finland, Lithuania, Czech Republic and the UK, in which the cartb
customer requirements. The transportation costs are given in Table 2.

The customers are located across northern Europe. Operations department of P&P has grouped customers according to
and produced estimates on the average transportation costs of serving each customer group from each warehouse (Tabl
in Finland to the warehouse in Finland has a limited capacity of 1100 tons per month. The Czeck warehouse has a limited
tons in the coming months due to maintenance.

The Sales & Marketing department has just released its demand estimates for the following month (Table 4).

Director of Sales & Operations has tasked you to develop analytics that support decision making on the distribution of ca
discussions you are able to convince the director of the added benefits of developing and deploying a network linear pro
these decisions.

transportation costs while satisfying the demands. Use decision variables 𝑥_𝑖𝑗 and transportation costs 𝑐_𝑖𝑗, where 𝑖,𝑗
a) Formulate mathematically a transhipment model that identifies the production quantities and the transportation volu

min ∑_(𝑖=1)^4▒ 〖∑ _(𝑗=4)^7▒𝑐_𝑖𝑗 𝑥 〗 _𝑖𝑗 + ∑_(𝑖=4)^7▒ 〖∑ _(𝑗=8)^20▒𝑐_𝑖𝑗 𝑥 〗 _𝑖𝑗

Constraints:

Finland: ∑_(𝑗=4)^7▒𝑥_𝑖𝑗 <= 12000 (i=1)

(1) Production capacities

Sweden: ∑_(𝑗=4)^7▒𝑥_𝑖𝑗 <= 2600 (i=2)

Belgium: ∑_(𝑗=4)^7▒𝑥_𝑖𝑗 <= 5100 (i=3)

∑_(𝑖=1)^3▒𝑥_𝑖𝑗 = ∑_(𝑗=8)^20▒𝑥_𝑖𝑗
(2) Transshipment nodes

∑_(𝑖=4)^7▒𝑥_𝑖𝑗 =𝐷_𝑗 , 𝑗∈{8,…, 20}

(3) Demand

𝑥_14≤1100
(4) Finland railway capacity

∑_(𝑖=1)^3▒𝑥_𝑖𝑗 <=5000 , j=6

(5) Czech warehouse capacity

Non negativity: 𝑥_𝑖𝑗>= 0

b) Implement transhipment model using spreadsheets and report the optimal total transportation cost (3 pts):
optimal objective function value: 3460050
c) To present the optimal result to the director prepare a visualization of the network using the attached map. To keep th
only those arcs through which there is a non-zero flow. Remember to visualize both flows from the mills to the warehous
to the customers. (1pt)

Table 2 Table 1
Mill
Tpt. cost per ton Production
1. Finland 2. Sweden 3. Belgium capacities
(tons)
 Production
capacities
(tons)
4. Finland 47.7 54.0 189.0
5. Lithuania 85.5 108.0 171.0
Warehouse
6. Czech 122.4 118.8 80.1
7. UK 144.0 126.0 63.9
Table 4
Table 3 Demand
Tpt. cost per ton
Warehouse
4. Finland 5. Lithuania 6. Czech 7. UK
8. Finland 90 180 252 288
9. Sweden 108 216 234 252

Customer areas
10. Norway 135 252 288 180
11. Iceland 144 279 360 117
Customer areas

12. Denmark 270 252 126 117

13. UK 279 360 287.1 63
14. Ireland 306 396 261 90
15. France 441 369 216 81
16. Benelux 387 342 207 90
17. Germany 324 171 72 171
18. Poland 252 72 54 333
19. Czech 315 135 54 288
20. Baltic 180 81 144 351
 Total 36 points)
c) Answer here
y production capacities are given in Table 1. From the mills
c and the UK, in which the cartboard is sheeted according to

ouped customers according to their geographical location

oup from each warehouse (Table 3). The rail line from the Mill
Czeck warehouse has a limited sheeting capacity of 5000

ng month (Table 4).

making on the distribution of cartboard. After some

deploying a network linear programming model to support

portation costs 𝑐_𝑖𝑗, where 𝑖,𝑗∈{1,…, 20} (3 pts)

ties and the transportation volumes which minimize total

_𝑖𝑗

portation cost (3 pts):

g the attached map. To keep the visualization clear, include

from the mills to the warehouses and from the warehouses

Grading
a) Is the formulation reasonable? (+0-3 pts)
Formulation is reasonable and correct. (Note that multiple
formulations exist.) (3 pts)
Formulation is reasonable, but there are minor mistakes (e
binary constraints missing. (2pts)
Formulation is reasonable, but there are major mistakes. (1
Mill
Completely unreasonable or no mathematical formulation
1. Finland 2. Sweden 3. Belgium
b) Model is correct, correct optimal solution is reported an
clearly presented (decision varibles, data, and formulas cle
not hidden in formulas but presented in data cells correspo
function and constraint coefficients). (3pts)
Small error in the model or in reporting solutions (-1 pt)
Model not clearly presented (-1 pt)
 Formulation is reasonable, but there are minor mistakes (e
binary constraints missing. (2pts)
Formulation is reasonable, but there are major mistakes. (1
Completely unreasonable or no mathematical formulation

12,000 2,600 5,100

b) Model is correct, correct optimal solution is reported an
clearly presented (decision varibles, data, and formulas cle
not hidden in formulas but presented in data cells correspo
function and constraint coefficients). (3pts)
Small error in the model or in reporting solutions (-1 pt)
Model not clearly presented (-1 pt)
There are major errors in the model (-2 pts)
Demand Model is poorly presented (-2 pts)
8. Finland 780 No spreadsheet implementation given (0 pts)
Not a linear model (0 pts)
9. Sweden 312
10. Norway 780
11. Iceland 156 C) Visualization includes all arcs through which there is non
12. Denmark 936
13. UK 3000
14. Ireland 1092
15. France 3744
16. Benelux 1092
17. Germany 4368
18. Poland 1248
19. Czech 780
20. Baltic 936
wer here

sonable? (+0-3 pts)

e and correct. (Note that multiple equivalent
s)
e, but there are minor mistakes (e.g., positivity or
ng. (2pts)
e, but there are major mistakes. (1pts)
e or no mathematical formulation given. (0 pts)

ect optimal solution is reported and the model is

on varibles, data, and formulas clearly marked; data
ut presented in data cells corresponding to objective
coefficients). (3pts)
or in reporting solutions (-1 pt)
nted (-1 pt)
 e, but there are minor mistakes (e.g., positivity or
ng. (2pts)
e, but there are major mistakes. (1pts)
e or no mathematical formulation given. (0 pts)

ect optimal solution is reported and the model is

on varibles, data, and formulas clearly marked; data
ut presented in data cells corresponding to objective
coefficients). (3pts)
or in reporting solutions (-1 pt)
nted (-1 pt)
n the model (-2 pts)
ed (-2 pts)
entation given (0 pts)
s)

all arcs through which there is non-zero flow (1pt)

 Spreadsheet implementation here (b)

Decision variables for production quantities from mills

(columns) to warehouse (rows)
destinations/
mills/warehouses 1. Finland 2. Sweden 3. Belgium warehouses
4. Finland 1100 928 0 4. Finland
5. Lithuania 2332 0 0 5. Lithuania
6. Czech 5000 0 0 6. Czech
7. UK 3092 1672 5100 7. UK

Costs 1. Finland 2. Sweden 3. Belgium Total cost Costs

4. Finland 47.7 54.0 189.0 102582 4. Finland
5. Lithuania 85.5 108.0 171.0 199386 5. Lithuania
6. Czech 122.4 118.8 80.1 612000 6. Czech
7. UK 144.0 126.0 63.9 981810 7. UK
1895778

constraints
capacity
finland 11524 <= 12,000
sweden 2600 <= 2,600
belgium 5100 <= 5,100

nodes
warehouse customer
4. Finland 2028 = 2028
5. Lithuania 2332 = 2332
6. Czech 5000 = 5000
7. UK 9864 = 9864

destination

8. Finland 780 = 780

9. Sweden 312 = 312
10. Norway 780 = 780
11. Iceland 156 = 156
12. Denmark 936 = 936
13. UK 3000 = 3000
14. Ireland 1092 = 1092
15. France 3744 = 3744
16. Benelux 1092 = 1092
17. Germany 4368 = 4368
18. Poland 1248 = 1248
19. Czech 780 = 780
20. Baltic 936 = 936

finland railway 1100 <= 1100

czech warehouse 5000 <= 5000

 Decision variables for transportation volumes from warehouse (rows) to destinations (columns)

8. Finland 9. Sweden 10. Norway 11. Iceland 12. Denmark 13. UK 14. Ireland 15. France 16. Benelux 17. Germany
780 312 780 156 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 4368
0 0 0 0 936 3000 1092 3744 1092 0

8. Finland 9. Sweden 10. Norway 11. Iceland 12. Denmark 13. UK 14. Ireland 15. France 16. Benelux 17. Germany
90 108 135 144 270 279 306 441 387 324
180 216 252 279 252 360 396 369 342 171
252 234 288 360 126 287.1 261 216 207 72
288 252 180 117 117 63 90 81 90 171
nations (columns)

18. Poland 19. Czech 20. Baltic

0 0 0
1248 148 936
0 632 0
0 0 0

18. Poland 19. Czech 20. Baltic Total cost

252 315 180 231660
72 135 81 185652
54 54 144 348624
333 288 351 798336
1564272
objective: min 3460050
ISM-C1004 - Business Analytics 1 - Assignment 2 (Total 36 points)

Project portfolio selection (6pts)

The Järviniemi area, located in the city of Niemelä in southern Finland, is regionally and nationally important nature p
variety of forests, peatlands, rock outcrops, lakes and small waters. Of the peatlands, which cover one-fifth of the are
forestry 50–70 years ago, which has resulted in a decrease in their biodiversity. In 2020, the City of Niemelä develope
important part of the plan was to make recommendations for the rewetting of the area’s forestry drained peatlands o
which peatlands should be rewetted first and by what measures to maximize the biodiversity benefits and to minimiz
(GHG) emissions. Peatlands can be a source or sink of GHGs, depending on how they are managed. An important asp
climate change, which in peatlands may be in trade-off with biodiversity protection.

Your task is to help the city of Niemelä in deciding the optimal combination of rewetting actions for each of the 20 pe
each stand there are three mutually exclusive options: do nothing, restoration and damming. Table 1 lists the change
actions. GHG emissions (CO2, methane, nitrous oxide) are measured in CO2 equivalent tons compared to no action o
predicted long-term (100 years) change in the number of plant species for which the habitat is favorable compared to
implementing the different actions.

A budget of 40 keuros is reserved for the rewetting actions. Moreover, stands 6 and 12 are next to each other and if t
then there will be a 3 keuro reduction in the total cost.
The city organized a half-a-day workshop to discuss the trade-off between GHG emission and biodiversity. Most of th
the City of Niemelä who were responsible for developing the management plan for the recreational area. In addition
Economic Development, Transport and the Environment, the Association for Nature Conservation and the company r
area. In this workshop the consensus was that a 2.1 ton reduction in the GHG emissions is equally preferred to a unit

GHG emissions and maximize biodiversity. Use the binary decision variables 𝑥_𝑗∈{0,1} and 〖𝑥′〗 _𝑗∈{0,1} to ind
a) Formulate mathematically a binary linear programming (BLP) model to find the optimal actions for each stand with

stand, respectively. Use 𝑠∈{0,1} as the decision variable required for modelling the cost synergy (3 pts).

𝑚𝑖𝑛 ∑_(𝑗=1)^20▒ 〖 (𝑥_𝑗 "(" 𝑔_𝑗 " + 𝑤" 𝑏_𝑗 ") + " 〖𝑥′〗 _𝑗 "(" 〖𝑔′〗 _𝑗 " " +"𝑤" 〖𝑏′〗 _𝑗)) 〗 _
constraints

∑_(𝑗=1)^20▒ 〖 (𝑥_𝑗 "+ " 〖𝑥′〗 _𝑗 "∗" 𝑐_𝑗) 〗 _ -3000s<= 40000

(1) budget

𝑥_𝑗 + 〖𝑥′〗 _𝑗 ≤1
(2) mutually exclusive

𝑥_6 + 𝑥_12−2𝑠≥ 0
(3) Synergy
𝑥_6 + 𝑥_12−2𝑠≤1
Binary: 〖𝑥′〗 _𝑗∈{0,1}, 𝑥_𝑗∈{0,1} , 𝑠∈{0,1}
b) Implement the BLP model using spreadsheets. Find the optimal portfolio of actions and report the reduction in GH
the model so that decision variables are on columns and constraints on rows. (3pts)

Optimal portfolio of actions:

Restoring: stands 2,7,16,17,18,19
Damming: stands 1,5,6,8,10,11,12,14
Do nothing: stands 3,4,9,13,15,20
Reduction in GHG emission: 298.86
Increase in biodiversity: 152

Table 1
Peatland Change in GHG emissions Change in biodiversity Cost (euros)
stand
Restoration gj Damming gj' Restoration bj Damming bj' Restoration cj
1 17.54 -39.37 25 4 4975
2 33.23 2.77 33 1 2807
3 102.51 25.65 30 2 4976
4 86.61 7.22 29 2 7367
 5 27.95 -62.73 24 -1 7926
6 31.19 -70.01 19 0 8848
7 11.03 -24.75 24 -1 3128
8 15.91 -35.70 27 -1 5154
9 116.77 29.22 37 1 5499
10 29.92 -67.16 22 -3 8642
11 23.42 -52.57 24 -1 6643
12 19.01 -42.66 21 -4 5392

13 77.85 6.49 30 1 6260

14 18.99 -42.63 27 -1 5386
15 274.23 68.62 22 3 12916
16 11.29 -25.35 23 0 3203
17 7.04 -15.79 27 -1 1997
18 39.58 9.90 33 1 1865
19 11.80 -26.49 19 0 3349
20 261.40 65.41 27 2 12312
2 (Total 36 points)

ly and nationally important nature protection and recreational area. It consists of a wide
nds, which cover one-fifth of the area, approximately 100 hectares were drained for
2020, the City of Niemelä developed a management plan for the Järviniemi area. An
e area’s forestry drained peatlands over the next five years. The key questions were
biodiversity benefits and to minimize the potential problems caused by greenhouse gas
hey are managed. An important aspect in the management plan was the mitigation of
on.

wetting actions for each of the 20 peatland stands in the Järviniemi area (Table 1). For
d damming. Table 1 lists the changes in biodiversity and GHG emissions from these
alent tons compared to no action over a 60-year period. Biodiversity is measured by the
the habitat is favorable compared to no action. Table 1 also lists the costs of

nd 12 are next to each other and if the restoration action is implemented for both stands
Grading
mission and biodiversity. Most of the participants were from different departments of
or the recreational area. In addition, there were participants from the Centre for a) Is the formulation rea
re Conservation and the company responsible for the recreational infrastructure of the Formulation is reasonab
ssions is equally preferred to a unit increase in biodiversity. exist.) (3 pts)
Formulation is reasonab
∈{0,1} and 〖𝑥′〗 _𝑗∈{0,1} to indicate the restoration and the damming of the 𝑗th
optimal actions for each stand with the given budget when the objective is to minimize constraints missing. (2pt
Formulation is reasonab
the cost synergy (3 pts). Completely unreasonab
" +"𝑤" 〖𝑏′〗 _𝑗)) 〗 _ b) Model is correct, corr
presented (decision vari
formulas but presented
constraint coefficients).
Small error in the mode
Model not clearly prese
There are major errors i
Model is poorly present
No spreadsheet implem
Not a linear model (0 pt
tions and report the reduction in GHG emission and the increase in biodiversity. Build
ts)

Cost (euros)

Damming cj'
2488
1404
2488
3683
3963
4424
1564
2577
2750
4321
3321
2696

3130
2693
6458
1601
999
933
1675
6156
 cj
cj'

Grading
a) Is the formulation reasonable? (+0-3 pts)
Formulation is reasonable and correct. (Note that multiple equivalent formulations
exist.) (3 pts)
Formulation is reasonable, but there are minor mistakes (e.g., positivity or binary
constraints missing. (2pts)
Formulation is reasonable, but there are major mistakes. (1pts)
Completely unreasonable or no mathematical formulation given. (0 pts)

b) Model is correct, correct optimal solution is reported and the model is clearly
presented (decision varibles, data, and formulas clearly marked; data not hidden in
formulas but presented in data cells corresponding to objective function and
constraint coefficients). (3pts)
Small error in the model or in reporting solutions (-1 pt)
Model not clearly presented (-1 pt)
There are major errors in the model (-2 pts)
Model is poorly presented (-2 pts)
No spreadsheet implementation given (0 pts)
Not a linear model (0 pts)
Spreadsheet implementation here (b)

Coefficients
Decision variables restoring xj damming x'j do nothing Restoration gj
1 0 1 0 17.54
2 1 0 0 33.23
3 0 0 1 102.51
4 0 0 1 86.61
5 0 1 0 27.95
6 0 1 0 31.19
7 1 0 0 11.03
8 0 1 0 15.91
9 0 0 1 116.77
10 0 1 0 29.92
11 0 1 0 23.42
12 0 1 0 19.01
13 0 0 1 77.85
14 0 1 0 18.99
15 0 0 1 274.23
16 1 0 0 11.29
17 1 0 0 7.04
18 1 0 0 39.58
19 1 0 0 11.80
20 0 0 1 261.40
s 1
w -2.1

constraints
Budget 39832 <= 40000
Mutually exclusive restoring xj damming x'j do nothing
1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=

1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=
 1 1 1 1=
1 1 1 1=
1 1 1 1=
1 1 1 1=
Synergy
x6 1
x12 1
2s 2
condition 1 0 >= 0
condition 2 0 <= 1
Damming gj' Restoration bj Damming bj' Restoration cj Damming cj'
-39.37 25 4 4975 2488
2.77 33 1 2807 1404
25.65 30 2 4976 2488
7.22 29 2 7367 3683
-62.73 24 -1 7926 3963
-70.01 19 0 8848 4424
-24.75 24 -1 3128 1564
-35.70 27 -1 5154 2577
29.22 37 1 5499 2750
-67.16 22 -3 8642 4321
-52.57 24 -1 6643 3321
-42.66 21 -4 5392 2696
6.49 30 1 6260 3130
-42.63 27 -1 5386 2693
68.62 22 3 12916 6458
-25.35 23 0 3203 1601
-15.79 27 -1 1997 999
9.90 33 1 1865 933
-26.49 19 0 3349 1675
65.41 27 2 12312 6156

1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
xjgj 113.9664
xj'gj' -412.829
sum -298.8626

xjbj 159
xj'bj' -7
sum 152

objective: min -618.0626

xjcj 16349
xj'cj' 26483
ISM-C1004 - Business Analytics 1 - Assignment 2 (Total 36 points)
Covering problem (6 pts)
A specific region (maakunta) in Finland needs to decide on the locations of healthcare clinics to serve its
municipalities (labelled A-G for confidentiality). Each clinic can serve all those municipalities that are
within a 23 minute driving distance. The travel times between the municipalities are given in Table 1.
Moreover, the cost of constructing a clinic varies across the municipalities. Projections of these costs are
given in Table 2.

a) Mathematically formulate a BLP model that identifies those municipalities to which clinics should be

minute driving distance. Use decision variables 𝑥_𝐴,…, 𝑥_𝐺∈{0,1}. (3 pts)

built so that the total construction costs are minimized and all municipalities have a clinic within a 23

min⁡〖 3.15𝑥_𝐴+ 〗 3.6𝑥_𝐵+3.15𝑥_𝐶+4.05𝑥_𝐷+2.85𝑥_𝐸+2.7𝑥_𝐹+3.6𝑥_𝐺

A and around A: 𝑥_𝐴+𝑥_𝐵+𝑥_𝐶≥1

Contraints:

B and around B: 𝑥_𝐴+𝑥_𝐵+𝑥_𝐷≥1

C and around C: 𝑥_𝐴+𝑥_𝐶+𝑥_𝐷≥1
D and around D: 𝑥_𝐵+𝑥_𝐶+𝑥_𝐷+𝑥_𝐹 "+" 𝑥_𝐺≥1
E and around E: 𝑥_𝐸 "+" 𝑥_𝐹≥1
F and around F:𝑥_𝐷+𝑥_𝐸 "+" 𝑥_𝐹≥1
G and around G: 𝑥_𝐷 "+" 𝑥_𝐺≥1
Binary: 𝑥_𝐴,…, 𝑥_𝐺∈{0,1}
b) Implement a spreadsheet model to solve this BLP and report the municipalities to which clinics should
be built as well as the optimal total cost. Build the model so that decision variables are on columns and
constraints on rows. (3 pts)
Municipalities to which cliniccs should be built: A, F, G
Optimal total cost: 9.45

Table 2. Construction
Table 1. Driving times between municipalities (minutes) costs (Meuros)
B C D E F G A
A 15 20 35 35 45 40 B
B 35 20 35 40 40 C
C 15 50 45 25 D
D 35 20 20 E
E 15 40 F
F 35 G
(Total 36 points)

linics to serve its

lities that are
ven in Table 1.
of these costs are

clinics should be
nic within a 23

hich clinics should

on columns and

Grading
Table 2. Construction
costs (Meuros) a) Is the formulation reasonable? (+0-3 pts)
Formulation is reasonable and correct. (Note that multiple equivalent formulations
3.15 exist.) (3 pts)
3.6 Formulation is reasonable, but there are minor mistakes (e.g., positivity or binary
3.15 constraints missing. (2pts)
4.05 Formulation is reasonable, but there are major mistakes. (1pts)
Completely unreasonable or no mathematical formulation given. (0 pts)
2.85
2.7 b) Model is correct, correct optimal solution is reported and the model is clearly
3.6 presented (decision varibles, data, and formulas clearly marked; data not hidden in
formulas but presented in data cells corresponding to objective function and
constraint coefficients). (3pts)
Small error in the model or in reporting solutions (-1 pt)
Model not clearly presented (-1 pt)
There are major errors in the model (-2 pts)
Model is poorly presented (-2 pts)
No spreadsheet implementation given (0 pts)
Not a linear model (0 pts)
 Spreadsheet implementation here (b)
Xa Xb Xc Xd Xe
Decision variables 1 0 0 0 0
Coefficients 3.15 3.6 3.15 4.05 2.85
Objective function min 9.45
constraints a 1 >= 1
b 1 >= 1
c 1 >= 1
d 2 >= 1
e 1 >= 1
f 1 >= 1
g 1 >= 1

lent formulations
itivity or binary

(0 pts)

model is clearly
ata not hidden in
nction and
Xf Xg
1 1
2.7 3.6
ISM-C1004 - Business Analytics 1 - Assignment 2 (Total 36 points)

Fixed charge (8 pts)

Take-A-Cape Analytics (TACA) is a consulting firm that specializes in developing analytics solutions for retail compa
has offices in Helsinki, Stockholm, Oslo and Vilna.

With the recently secured series B funding TACA is planning to open new offices in one or more European cities. T
cities includes Frankfurt, Warsaw, Paris, Amsterdam, Madrid, Milan and Vienna. The startup costs of the new offic

Each new office will require a specific number of analytics experts from the current offices as they have deep know
the company’s products (Table 1). The company has conducted a survey to identify the number of experts in each
transfer to a new office (Table 2). The costs of transferring an expert from one city to another - including housing c
of living adjustments to the salaries - are listed in Table 3.

that an office can be established in city with the rank 𝑗 only if offices are also established to all cities with ranks 1
The board of the company has ranked the new offices in terms of the potential they have to increase future sales

a) Formulate mathematically a MILP model that for a prespecified number of new offices 𝑛 determines (i) the citi

transfer) are minimized. Use the binary decision variables 𝑦_𝑗 and continuous decision variables 𝑥_𝑖𝑗, where 𝑗∈
should be established and (ii) how experts from the current offices are transferred to these new offices so that tot

(4pts)

𝑚𝑖𝑛∑_(𝑖=1)^4▒ 〖∑ _(𝑗=1)^7▒𝑡_𝑖𝑗 𝑥 〗 _𝑖𝑗 + 〖∑ _(𝑗=1)^4▒𝑐_𝑗 𝑦 〗 _𝑗

Constraints:

Frankfurt: ∑_(𝑖=1)^4▒𝑥_𝑖𝑗 _ = 9 (j=1)

(1) Number of expert required:

Warsaw:∑_(𝑖=1)^4▒𝑥_𝑖𝑗 =7 (j=2)
Milan: ∑_(𝑖=1)^4▒𝑥_𝑖𝑗 =14 (j=3)
Amsterdam: ∑_(𝑖=1)^4▒𝑥_𝑖𝑗 =12 (j=4)
Madrid: ∑_(𝑖=1)^4▒𝑥_𝑖𝑗 =11 (j=5)
Paris: ∑_(𝑖=1)^4▒𝑥_𝑖𝑗 =8 (j=6)
Vienna: ∑_(𝑖=1)^4▒𝑥_𝑖𝑗 =18 (j=7)

Helsinki: ∑_(𝑗=1)^7▒𝑥_𝑖𝑗 ≤24 (i=1)

(2) Number of expert available:

Stockholm: ∑_(𝑗=1)^7▒𝑥_𝑖𝑗 ≤19 (i=2)

Oslo: ∑_(𝑗=1)^7▒𝑥_𝑖𝑗 ≤16 (i=3)
Vienna: ∑_(𝑗=1)^7▒𝑥_𝑖𝑗 ≤21 (i=4)

𝑦_5, 𝑦_7>= 𝑦_1, 𝑦_2, 𝑦_3>= 𝑦_4, 𝑦_6

(3) Rank condition:

(4) 𝑦_𝑗 binary

(5) 𝑥_𝑖𝑗 integer
(6) maximum number of offices: ∑_(𝑗=1)^7▒𝑦_𝑗 =n
Non negativity. x,y>=0
b) The board decides the budget for the expansion to be 22 million euros. Implement the MILP model on the spre
determine the maximum number of offices that can be established and the cities in which the new offices should

constraints on rows (4 pts). [HINT: solve the model repeatedly for decreasing values on 𝑛 until the optimal cost do
cities and the minimal cost of establishing new offices to these cities. Build the model so that decision variables ar

Cities: Frankfurt, Madrid, Vienna

Maximum number of offices: 3
Minimal cost: 1991220
Table 1 Table 2
Number of Experts
Candidate city for new Startup Cost
experts Current office (i) available
office (j) sj (Meuros)
required rj ai
1. Frankfurt 3.135 9 1. Helsinki 24
2. Warsaw 4.95 7 2. Stockholm 19
3. Milan 6.6 14 3. Oslo 16
4. Amsterdam 4.62 12 4. Vilna 21
5. Madrid 5.61 11
6. Paris 3.795 8
7. Vienna 7.425 18

Table 4

Candidate city for

Rank
new office (j)

1. Frankfurt 2
2. Warsaw 2
3. Milan 2
4. Amsterdam 3
5. Madrid 1
6. Paris 3
7. Vienna 1
 otal 36 points)

tics solutions for retail companies. The company currently

ne or more European cities. The shortlist of potential

startup costs of the new offices are given in Table 1.

offices as they have deep knowledge and understanding of

he number of experts in each current office willing to
another - including housing costs, moving costs and cost

shed to all cities with ranks 1, …,𝑗−1.

have to increase future sales (Table 4). The board states
Grading
ffices 𝑛 determines (i) the cities in which the new offices
a) Is the formulation reasonable? (+0-4 pts)
Formulation is reasonable and correct. (Note that multiple equ
ion variables 𝑥_𝑖𝑗, where 𝑗∈{1,…,7} and 𝑖∈{1, …,4}.
these new offices so that total costs (i.e., startup and formulations exist.) (4 pts)
Formulation is reasonable, but there are minor mistakes (e.g.,
binary constraints missing. (3pts)
Formulation is reasonable, but there are major mistakes. (1pts
Completely unreasonable or no mathematical formulation give

b) Model is correct, correct optimal solution is reported and th

clearly presented (decision varibles, data, and formulas clearly
not hidden in formulas but presented in data cells correspond
objective function and constraint coefficients). (4pts)
Small error in the model or in reporting solutions (-1 pt)
Model not clearly presented (-1 pt)
There are major errors in the model (-2 or -3 pts)
Model is poorly presented (-2 or -3 pts)
No spreadsheet implementation given (0 pts)
Not a linear model (0 pts)

nt the MILP model on the spreadsheet and use it to

which the new offices should be established. Report the

on 𝑛 until the optimal cost does not exceed the budget]

l so that decision variables are on columns and
Table 3

Transfer costs 1. 2. 3. 5. 7.
4. Amsterdam 6. Paris
tij (keuros) Frankfurt Warsaw Milan Madrid Vienna

1. Helsinki 34.65 49.5 57.75 24.75 41.25 24.75 74.25

2. Stockholm 24.75 56.1 85.8 51.15 64.35 33 95.7
3. Oslo 29.7 47.85 70.95 41.25 56.1 34.65 87.45
4. Vilna 39.6 37.95 47.85 33 51.15 42.9 77.55
 Spreadsheet implementation her
1. Frankfurt
Decision variables for
whether should start
up an office in cities y1
1
startup cost 3.135
objective: min 16.17

ble? (+0-4 pts)

d correct. (Note that multiple equivalent

t there are minor mistakes (e.g., positivity or

pts)
t there are major mistakes. (1pts)
no mathematical formulation given. (0 pts)

ptimal solution is reported and the model is

ribles, data, and formulas clearly marked; data
esented in data cells corresponding to
aint coefficients). (4pts) constraints
reporting solutions (-1 pt)
-1 pt) number of expert required
model (-2 or -3 pts) 1. Frankfurt 9
or -3 pts) 2. Warsaw 7
on given (0 pts)
3. Milan 14
4. Amsterdam 12
5. Madrid 11
6. Paris 8
7. Vienna 18
expert available
1. Helsinki 24
2. Stockholm 18
3. Oslo 16
4. Vienna 21
rank y1
1
1

-1
 -1

maximum number of offices

3=
preadsheet implementation here (b)
2. Warsaw 3. Milan 4. Amsterdam 5. Madrid 6. Paris 7. Vienna
Decision variables
for number of
expert transfer
y2 y3 y4 y5 y6 y7 from
0 0 0 1 0 1 1. Frankfurt
4.95 6.6 4.62 5.61 3.795 7.425 2. Warsaw
3. Milan
4. Amsterdam
5. Madrid
6. Paris
7. Vienna

Tranfer costs
1. Frankfurt
2. Warsaw
3. Milan
4. Amsterdam
5. Madrid
6. Paris
7. Vienna
Objective: min

= 9
= 7
= 14
= 12
= 11
= 8
= 18

<= 24
<= 19
<= 16
<= 21
y2 y3 y4 y5 y6 y7
-1 1 >=
-1 1 >=
1 -1 0 >=
1 -1 0 >=
1 -1 0 >=
1 -1 0 >=
1 0 >=
-1 1 1 >=
-1 1 1 >=
-1 1 1 >=
1 -1 1 >=
 1 0 >=
-1 1 1 >=
-1 1 1 >=

-1 1 1 >=
-1 1 1 >=

3
 Helsinki Stockholm Oslo Vienna

x1 x2 x3 x4
0 9 0 0
0 0 0 7
0 0 0 14
12 0 0 0
11 0 0 0
0 8 0 0
1 1 16 0

Helsinki Stockholm Oslo Vienna Total cost

34.65 24.75 29.7 39.6 222.75
49.5 56.1 47.85 37.95 265.65
57.75 85.8 70.95 47.85 669.9
24.75 51.15 41.25 33 297
41.25 64.35 56.1 51.15 453.75
24.75 33 34.65 42.9 264
74.25 95.7 87.45 77.55 1569.15
3742.2 19912200

0
0
0
0
0
0
0
0
0
0
0
0
0
0

0
0