W15_LUCKY_Decision to repair
or to trade in a car using Decision Tree
1.
Problem
Recognition
I have been asked for
advice by a colleague. He has an old Hyundai saloon car valued at $1,592. He
has a budget of $600 for the project. An
experienced Honda mechanic, a close associate, has offered to refurbish the car
at an estimated cost of $250 in one week, hopefully, before my colleague
travels to Ghana. A certified OEM (Hyundai) Centre offered to revamp the car at
a cost of $350 to make it look like new. The Centre also offered to accept a
trade in of the car at a cost of $2,290.
2.
Problem Definition
A set of quick calculation
reveals there is problem. My colleague has a budget of $600.
Honda mechanic repair cost -
$250
OEM repair cost - $350
Second hand value of car -
$1,592
Trade-in car option cost -
$2,290
Given his budget, which
alternative will be the best? This is a replacement problem that I would like
to solve using the Decision Tree Tool and Technique.
a.
Assumptions
Key assumptions include:
·
The second hand value estimate of the car is reliable with 90 percent probability
·
The Honda
mechanic estimate is good, and will meet deadline of one week with 80 per cent
probability
·
The OEM
Centre estimates are correct and cast in stone with 100 percent probability
·
The OEM
will be willing to buy the car upon seeing it and replacement car available
with 90 percent probability
3.
Feasible
Alternatives
Feasible alternatives
include:
A.
Do nothing –
abandon the idea altogether
B.
Engage the
Honda mechanic
C.
Engage
the OEM Centre
D.
Trade in
the car with the OEM Centre
In an earlier blog,
W7_Lucky_Decision to repair or to trade in a car, the proposed feasible
alternative arrived at was alternative #3 above. Suppose the replacement car is available at
the OEM Centre for the one week period. As each day passes a decision point is
reached until the replacement is finally consummated. In this case, the problem
has changed into a replacement problem. There is no chance of
time-value-of-money due to the short duration of the envisaged transaction –
one week- at this instance. Thus I would like to solve it using the Decision
Tree Tool and Technique.
4.
Development of outcomes for each alternative
The decision tree outlining
the scenarios is shown in figure 1 below. On this tree there is one decision
point:
The outcomes for each
alternative are as follows:
1.
Do
nothing or
2.
Fix the
car
Along the way to the ultimate
outcomes is one event node, node X from which I can chose any of the following:
1.
Engage the
Honda mechanic
2.
Engage
the OEM Centre
3.
Trade in
the car with the OEM Centre
Taking each ultimate outcome
in turn, the following calculations result:
Outcome A: For the ‘Don’t fix
the car’ option, the outcome is zero
Outcome B: If the Honda
mechanic is engaged,
$500 (10%) = 50
$300 (20%) = 60
$250 (70%) = 175
Value = $285
Outcome C: If the OEM Centre
is engaged,
$500 (10%) = 50
$450 (20%) = 90
$350 (70%) = 245
Value = $385
Outcome D: If I advise
trade-in the car,
$3000 (10%) = 300
$2500 (10%) = 250
$2290 (80%) = 1832
Value = $2382
4. Selection Criteria
The selection criterion is
the expected value that is LOWEST and assures
1.
On time
delivery
2.
Quality
standard
3.
Cost (within
budget)
5
Analysis and Comparison of the alternatives
Working back through the
paths:
The
Expected value EV for activity ‘Don’t fix the car’ is $0
Net outcome A: Zero
The
Expected value EV for activity ‘Engage Honda mechanic’ is $285
Net outcome B: $285
Percent Probability = 30%
The
Expected value EV for activity ‘Engage OEM Centre’ is $385
Net outcome C: $385
Percent Probability = 60%
The
Expected value EV for activity ‘Trade-in Car’ is $2382
Net outcome D: $2382
Percent Probability = 10%
·
Monetary
consideration
To stay
within budget, alternatives A, B and C are the only acceptable feasible
alternatives, with alternative C being the best option due to its high
probability.
·
Introducing risk into the solution
This applies to all three alternatives.
i.
Failure
for on-time delivery – Travel to Ghana in public car [Best option is
Alternative D]
ii.
Quality
issue – Instigate rework; make alternative travel arrangement [Best options are
alternatives C and D]
iii.
Cost
changes – Request for firm price from service providers [All three options are
acceptable alternatives]
·
Nonmonetary consideration
These include schedule and quality standard.
If schedule is considered, alternative D becomes the
only feasible solution.
If quality is considered, alternatives C and D become
better feasible alternatives.
6.
Selection of preferred alternative
I am prepared to advice my
colleague to choose alternative C as my preferred alternative. He is would save
$215 to cover any quality or schedule risk.
7.
Performance monitoring and post evaluation of results
The car shall be driven
within town on test drives. Hopefully all performance issues with the car will
surface and be resolved before being driven to Ghana.
Reference
1.
Sullivan,
W., Wicks, E., Koelling, P., Kumar, p., & Kumar, N. (2012).Chapter 12
Probabilistic Risk Analysis (pp. 528 - 533). Engineering economy (15th edition). England: Pearson Education Limited.
2.
Brassard, M. & Ritter D.
(2010). Chapter 24 Tree Diagram (pp.200 – 209). The Memory Jogger 2 (2nd edition). USA:GOAL/QPC
3. Giammalvo,
P. (2012, October 22). Integrated
portfolio (asset), program (operations) and project management methodology
course (cost engineering) slides (An AACE methodology course). Lagos,
Nigeria: Lonadek
AWESOME, Lucky!!! Nice case study!!!
ReplyDeleteGlad to see you catching back up again and leading by example!! I was getting worried about you!!
BR,
Dr. PDG, Jakarta
Dr. PDG,
ReplyDeleteI was worried myself due to pressure of work following a reorganization at my place of work! I will try to keep up. I have a lot to do on problem solviing!
Regards,
Lucky
To catch up the HPE800 badge, the McLaren MP4-12C has been upgraded by the finest Texas tuners in town – making the car kicks its horsepower up to 800hp. In my point of view, I think the reason why Hennessey didn't lay their hands on suspension upgrade is because they feel that the OEM parts are enough to handle the added viciousness of the engine.
ReplyDeleteAndre Brennan