3. Issue
What issue are we modelling?
Skyride
Top
Skyride
Base
Dragon Trail
Jungle Trail
688m
628m
New Joiners
Re-joiners
Issue: Long queueing time
1 Widen Track
2 Create a third trail
Parameter:
1 Peak Hour: 10.30am – 12.30pm
2 Weekend
6. Inflow: New Joiners
Simulating the number of new joiners flowing into the system
Assumptions:
- The inter-arrival time between visitors follows an
exponential distribution
- The average inter-arrival time is 0.21mins
- Ushers will let a batch of new joiners in every 1.5mins
Inter-arrival Time = – Average * LN (1-Rand())
Formula:
Illustration:
Inter-arrival
Time
Arrival Time Time Intervals # Time New Joiners
0 1 0
0.550853514 0.550853514 2 0.3
0.012227288 0.563080802 3 0.6
0.001761651 0.564842454 4 0.9
0.013120036 0.577962489 5 1.2
0.049912837 0.627875326 6 1.5 10
0.228993028 0.856868354 7 1.8
0.322342296 1.17921065 8 2.1
0.096762228 1.275972878 9 2.4
0.138550518 1.414523396 10 2.7
0.040588191 1.455111587 11 3.0 7
0.283460628 1.738572215 12 3.3
<1.5
COUNTIF
7. J is a longer trail so
the luges arrive later
than D
Inflow: Rejoiners
Simulating the number of rejoiners flowing into the system
Assumptions:
- Riders’ decision to rejoin follows a Binomial distribution
- The probability of rejoin is 0.6
- Interval between each luge release is 1.8mins
Formula:
Illustration:
Rejoins from each trail = CRITBINOM (# riders finishing trails, 0.6, Rand())
# riders
finishing J
Rejoin from J
# riders
finishing D
Rejoin from D
1 0
3 0
5 4
5 3
1 0
5 2
1.8mins
1.8mins
Total Rejoin (t)
= SUM (Rejoin frm J & D at t)
8. Skyride: Queue Length
Simulating the Queue Length for the Skyride
Formula:
Queue
Length (t)
Queue
Length (t - 1)
# Skyriders
leaving (t – 1)= + -Total Inflow
(t)
New Joiners
+
Rejoiners
Queue
Time =
Queue
Length (t)
Ave. #
Skyriders
leaving with
each chairlift
x
Frequency of
chairlift:
0.3mins
Skyriders can
choose to sit in
groups of 1, 2, 3 or
4 on the chairlift.
How to
simulate??
9. Skyride: Riders per Chairlift
Simulating the number of skyriders on each chairlift
Assumptions:
- 1 Chairlift leaves every 0.3mins
- Maximum no. of riders on each chairlift – 4 pax
- It is equal likely for riders to board in groups of 1,2,3 or 4
- Chairlift takes 5.1mins to travel from base to peak
Formula:
If Queue Length < 4
RANDBETWEEN (1, Queue
Length)
# Skyriders
leaving (t ) =
If Queue Length > 4
RANDBETWEEN (1,4)
# Riders arriving at peak
and choosing which
luge trail to take
5.1mins later…
How many riders would
choose Jungle trail /
Dragon trail???
10. Luge: Queues Length
Simulating the Queue Length for the Luge Trails
Assumptions:
- Riders are equally likely to choose Jungle or Dragon trail
- Every 0.3mins, one chairlift will carry a max of 4 to peak
- Let “1” represent Jungle and “2” represent Dragon
Formula:
# Luge Riders
Choice of
1st Rider
Choice of
2nd Rider
Choice of
3rd Rider
Choice of
4th Rider
Choose J Choose D
X
Those
arriving on
the chairlift
Rider 1: IF(X>0, RANDBETWEEN(1,2),0)
Rider 2: IF(X>1, RANDBETWEEN(1,2),0)
Rider 3: IF(X>2, RANDBETWEEN(1,2),0)
Rider 4: IF(X>3, RANDBETWEEN(1,2),0)
J: Count if the choice is “1”
OR
D: Count if the choice is “2”
# Luge
Riders
Arriving
Choice of
1st Rdier
Choice of
2nd Rdier
Choice of
3rd Rdier
Choice of
4th Rdier
Choose
Jungle
Choose
Dragon
0 0 0 0 0 0 0
4 1 2 1 1 3 1
2 2 2 0 0 0 2
2 1 1 0 0 2 0
Illustration:
11. Formula:
Queue
Length for
J (t)
Queue
Length for J
(t - 1)
# riders released
(t – 1)
= + -# Riders
Chose J (t)
Luge: Queues Length (ctd’)
Simulating the Queue Length for the Luge Trails
# Luge
Riders
Arriving
Choice of
1st Rdier
Choice of
2nd Rdier
Choice of
3rd Rdier
Choice of
4th Rdier
Choose
Jungle
Choose
Dragon
0 0 0 0 0 0 0
4 1 2 1 1 3 1
2 2 2 0 0 0 2
2 1 1 0 0 2 0
Max 5 Riders will be
released in batches
every 1.8mins, may be
less than 5…
How to simulate??
12. Outflow: Leave/Rejoin
Simulating the Queue Length for the Skyride
Same as Rejoiner at Inflow
Cycle REPEATS
(for 2-hours worth of data)
# riders
finishing J
Rejoin from J
# riders
finishing D
Rejoin from D
1 0
3 0
5 4
5 3
1 0
5 2
1.8mins
1.8mins
Recall:
If there are less than 5 people
in the queue, all the people will
be released.
If there are more than 5 people,
a max of 5 will be released
Assumptions:
- A max of 5 riders will be released every 1.8mins
15. Option 1 Results
What is the average queue time if we expand the trails?
If there are less than 6 people
in the queue, all the people will
be released.
Changed Assumption:
- A max of 6 riders will be released every 1.8mins
If there are more than 6
people, a max of 6 will be
released
1 2
Average
Queue Time
28.4 mins
26.1 1 2 3 4 5 6 7 8 9 10
1 28.4 26.7 33.1 32.9 24.6 25.9 33.7 27.6 28.5 26.5
2 28 26.5 28.6 25.4 30.7 34.8 29 26.3 26.7 24.6
3 27.5 30.6 30.6 23.85 21.1 31.5 32.8 28.3 24.5 25.7
4 25.8 32.2 27.5 28.5 24.9 27.7 29.2 28.5 31.4 28.5
5 28.2 23.9 27 33.8 28.9 26 27.7 30.1 27.5 26.1
6 29.1 27.2 37.2 30.4 32.3 28.4 32.2 25.5 25.05 27.9
7 26.1 28.4 31.8 29 25 33.1 33.8 35.4 23.7 25.3
8 26.7 30.7 27.4 25.4 28.7 28.6 29 29.6 27.4 36.1
9 31.8 28.9 28.4 25.4 29.5 29 24 22.5 35.4 27
10 26 28.4 35.8 29.9 25.6 19.2 28.7 27.8 32 24.7
18. Overall Results
Sensitivity Analysis and Summary of all Results
10
12
14
16
18
20
1.2 2.2 3.2 4.2 5.2
Ave.QueueTime
Luge Time
Sensitivity AnalysisOption 2 – Sensitivity Analysis
- Additional trail Long or Short?
- Long trail = takes a longer time to travel
down to base “Luge Time”
- CONCLUSION: The length of the additional
trail has little effect on the queue time
Scenario Average Queue Time (mins) % Improvement
Base Case 41.5 NA
Enlarged Trails 28.4 32%
Additional Trail – Shorter trail 17.9 57%
Additional Trail – Longer tail 18.5 55%
- Both options are effective in enhancing the queue time
- An additional trail is more effective
- A short or a long trail does not alter the results much
- Considering that a longer trail would cost more to construct, our suggestion to the
management is to build an additional shorter trail to improve the queue time.
Conclusion: