2. The wise choice
“Excellence is never an accident.
It is always the result of high
intention, sincere effort, and
intelligent execution; it represents
the wise choice of many
alternatives - choice, not chance,
determines your destiny.”
― Aristotle
3. 3
Outline
Forecasting.
Basic concept of the forecasting model
Model constrains
Max/min signal tracking approach
Accuracy forecasting Matrix
Setting Targets
K.P.I level for the airports
Case Study : Brussels Airport
Basic Data Base
Trend Analysis
Scenarios
Optimum Solution
Passengers Forecasting
Aircraft Movements Forecasting
Cargo Forecasting
Best Practice
Icao Monitoring Report
Passengers Forecasting
Aircraft Movements Forecasting
Cargo Forecasting
Last Year – Passengers
Last Year – Aircraft Cycles
Results
Passengers
Aircraft Movements
Cargo
Summary
Conclusion's
10. 1.5 Setting Targets
Most of airports in the world
working on a clear objectives and
that’s come with clear targets
which lead us to set a clear picture
of forecasting process.
Based on that, our objective is to
develop a clear massage for top
managements for the key
performance figures of the airline,
not just to compare month by
month approach but to develop the
right path ( time series ) in the
future to set the right targets which
consequently develop K.P. I
levels for the airports
12. 2. Case Study : ( Brussels Airport )
Brussels Airport (IATA: BRU, ICAO:
EBBR) (also called Brussel-Nationaal /
Bruxelles-National (Brussels-National) or
Zaventem) is an international airport
6.5 NM (12.0 km; 7.5 mi) northeast[2] of
Brussels, the capital of Belgium. In 2015,
more than 23 million passengers arrived
or departed at Brussels Airport, making it
the 21st busiest airport in Europe. It is
located partially in Zaventem, partially in
the Diegem area of Machelen,[3] and
partially in Steenokkerzeel, in the Flemish
Region of Belgium. It is home to around
260 companies, together directly
employing 20,000 people and serves as
the home base for Brussels Airlines,
Jetairfly and Thomas Cook Airlines
Belgium.
13. 2.1 Basic Data Base
( Three years data )
36 months data files
14. 2.1 Basic Data Base
( Three years data )
36 months data files
15. 2.2 Trend Analysis
Trend Analysis :
The concept of trend analysis is to study the impact of
displacement factor that will lead to minimize the error of
the last year data over Optimum Solution.
19. 2.3.1 First Scenario
First Scenario: Optimum Solution
This is the solution of minimum resulting errors, i.e the
deviations are uniformly distributed on both sides of the
trend line, we will find some errors here and some errors
there over all the period (three years) , but it is the best
solution to represents the data – by the highest R and min
bound of signal tracking.
23. 2.3.2 Second Scenario
Second Scenario: Best Practice
It reflects the impact of most recent data in the analysis
(Last Year), with the general trend model, this can be
done by adding a new constrains for the model for last
year data. The result, of that will be minimum deviations –
this reflects the best practice to compare the recent data
with same data of the same period of last year, the good
thing here is defining the general trend of the model that
reflects the last year data and used to forecast without
stacking in the past..
24. 2.3.2.1 ICAO Report
Second Scenario: Best Practice
Good example of using last year
data to evaluate the performance in
aviation industry is ICAO Monitoring
Report, as shown in the graph.
ICAO Monitoring
Report – Aug 2016
But with out any inference of the
trend just stacking in the past.
28. 2.3.2.4 Cargo Forecasting
Scenario Two - Best Practice
For Cargo Forecasting – Best Practice, this is not
Applicable, as there are a lot of discrepancies and high
deviations in last year Data.
But in general the
Model is Fair for the
first scenario, as R
square is 85% (greater
than 80%) and the
signal tracking is
bounded by ±4.5
( close to ±4)
29. 2.4 Last Year - Passengers
Optimum
Solution
Scenario
Best
Practice
Scenario
This slight change/ adjustment cause
2016 Pax forecast to drop by 1,178,043
30. 2.5 Last Year– A/C Movements
Optimum
Solution
Scenario
Best
Practice
Scenario
This slight change/ adjustment cause 2016
A/C movements forecast to drop by 7172 cycles
35. 4. Summary
First Scenario :
Optimum Solution
In this scenario we use a max/min
signal tracking approach, which
simply define the best possible
solution for the fitting data, so to
minimize the errors, it act to
distribute on both side of the trend
model, defining the general trend
of all data at the highest value of
R - square. This lead us to the
following results
R2 = 98.5 %
S.T.= 6.47 (2013-2015)
2016(F) = 26,669,778 Pax
2017(F) = 29,687,771 Pax
36. 4. Summary
Second Scenario :
Best Practice
Based on the previous
scenario , we add additional
constrains to force the model
to pass on the last year data,
the impact of that is a slight
change in R – square and
expand range in signal
tracking, but the result of
forecasting is significantly
change. This lead us to the
following result
R2 = 98.4 %
S.T.= 9.75 (2013-2015)
2016(F) = 25,491,735 Pax
2017(F) = 27,873,109 Pax
37. 5. Conclusions
First : The scenario that has lower value forecast will be the
fair one.
Second : Best Practice - Scenario
By setting the new constrains in the model, this will impact
the general trend of the previous scenario , we add
additional constrains to force the model to pass on the last
year data, the impact of that is a slight change in R – square
and expand range in signal tracking, but the result of
forecasting is significantly change. This lead us to the
following result
2016(F) = 25,491,735 Pax
2017(F) = 27,873,109 Pax
2016(F) = 247,958 Cycle
2017(F) = 258,525 Cycle