Cruise Liner Schedule
Jason Deleon Steve Rockwell
Will Wathen
Background
• Vacation Cruise Industry (2012)– Over 20 Million passengers – Generated over $42 Billion in economic activity (US)
• OR Consultants hired by Major Cruise Liner– Analyze current voyage network– Minimize cost– Itinerary recommendations– Analyze resiliency of network– Impacts of catalysts
*Paid hefty sum for analysis, no need to question resultsM
Real World Problem
• Network Design• Measures of Effectiveness on an itinerary for
an n day cruise– Cost (Transit and Port Fees)– Fun - Utility
• Analyze effects of desired fun– Relate to target demographics
• Impacts of catalyst i.e. weather, coup
M
Network Map Overview
M
Network Description
• Cruise Liner Voyage Network
– Consists of all possible voyage routes
– Connects all possible ports for a particular cruise liner (Royal Caribbean)
– Time-layered network • Each layer is a day
M
Network Design
• Nodes: Port on a given cruise day i.e Haiti 3– 16 Ports x 8 days = 128 nodes – S & T = Homeport = Fort Lauderdale, FL
• Node Data: Fun Factor, Lat/Long
• Edges: Connection of two ports from one time layer to another
– Bahamas 2 ------ Haiti 4 (Two Day Transit)
• Edge Data: Cost = Transit Cost + Port FeesM
Five Day Time Layer Network (Subset)
Day 0
Day 5
Day 4
Day 2
Day 3
Day 1
HA
BAH HA DR
DRHABAHFL
DR
FL
FL BAH
BAHFL
BAHFL
BAH HA DRFL
HA DR
HA DR
Five Day Time Layer Network (Subset)
Day 0
Day 5
Day 4
Day 2
Day 3
Day 1
HA
BAH HA DR
DRHABAHFL
DR
FL
FL BAH
BAHFL
BAHFL
BAH HA DRFL
HA DR
HA DR
Nodes
M
Start/End Node
M
Example of Route
M
Day 0
Day 1
Day 3
Day 2
Day 5
Mathematical Model
Shortest Path MILP• Minimizing Port Costs & Transit Costs • Netflow Constraints
• Design Constraints (Real World):• No overnight stays in port• Cannot return to visited port other than Homeport• Length of cruise in days
• AnalysisFun Factor constraint
M
Assumptions
• Direct Path - Navigable Route• Great Circle Distance• Average Speed of ship is constant : 25 knots• Max Range: 400 NM/day• Fun Factor: Excursions/Affordability• Port Fees: $10-$25 per person. • Haiti- $0 port fees - Long Term Lease in
Labadee (Royal Caribbean).
M
Situation Dependent Variables
• Travel Miles per day• Cruise Length in days• Cost of Fuel• Fuel Consumption rate• Ship speed • Port Fees• Fun Factor• Range/day* All Variables can be modified to correspond to real cruise network
M
Results: Min Cost
• 5-day Route:– FortLaud0, Bahamas1, Haiti3, FortLaud5– Cost: $425,194
• 7 -day Route: – FortLaud0, CaymanI2, Haiti4, Bahamas6, FortLaud7– Cost: $ 610,296
• 10-day Route:– FortLaud0, Bahamas1, StThomas4, PuertoRico5,
StMaarten6, Haiti8, FortLaud10– Cost: $ 891,990
M
Adding Utility
• Added Utility to the model– Fun Factor
• Utilized Constraint to ensure a certain level of “Fun”
• Increased Fun Factor from minimal feasible solution to maximum feasible solution
– Cost Vs. Fun Relationship– Demographic Comparisons
M
Utility: Fun Factor
• Cruise Length Fixed at 7 days• 16: FortLaud0, Bahamas1, Haiti3 , CaymanI5, FortLaud7
– Cost: $ 610,296 • 19: FortLaud0, CaymanI2 , DomRep4, Bahamas6, FortLaud7
– Cost: $ 721,276 • 21: FortLaud0, StThomas3, PuertoRico4, DomRep5, FortLaud7
– Cost: $ 751,934
• 22: FortLaud0, StMaarten3, StThomas4, DomRep5, FortLaud7 – Cost: $ 836,982
• 24: FortLaud0, DomRep2 , Aruba3 , CaymanI5, FortLaud7 – Cost: $ 902,282
M
Fun Factor vs. Cost
P
Target Market (<$75k)
- Marketing cruises based upon demographic.- Design cruise for family with combined income <$75k (~35% of passengers).
- Priority of customer may be price and not necessarily itinerary.
- Example: 16: FortLaud0, Bahamas1, Haiti3 , CaymanI5, FortLaud7
Cost: $ 610,296
Target Market (>$75k)
- Marketing cruises based upon demographic.- Design cruise for family with combined income >$75k (~65% of passengers).
- Priority of customer may be itinerary and not necessarily price.
- Example: - 24: FortLaud0, DomRep2 , Aruba3 , CaymanI5, FortLaud7
Cost: $ 902,282
Multi-Segment Strategy
- Utilizing a strategy to target multiple market segments that are based upon demographics.
- If only one itinerary is marketed, what would attract the most customers?
- Low price while providing moderate/high level of fun factor.
-Example:• 21: FortLaud0, StThomas3, PuertoRico4, DomRep5, FortLaud7
– Cost: $ 751,934
Attack
Attack
Scenario:• 7- day Itinerary for “most fun” cruise route• 2 days prior to departure
– Hurricane Thelma hits– Randomly takes out ports– Follow on attacks
• Consultants called to find alternate route with a near equivalent fun factor
Max Fun Route - No Attack
M
1 Attack
M
2 Attacks
M
3 Attacks
M
4 Attacks
M
5 Attacks
M
6 Attacks
M
7 Attacks
M
8 Attacks
M
9 Attacks
M
10 Attacks
M
Resiliency
Conclusion
• Shortest Path MILP– Minimizing Port Costs & Transit Costs– Analyzed impact of Fun Factor
• Positive Correlation (Fun & Cost)– Most Fun Cruise 1.5X as expensive as least Fun– Translates directly to Customer– Recommended Cruises based on Utility value
• Effects of Attacks & Resilience• Key Take-aways
P
Follow-on Work
-Enable User Interface -Multiple Cruise Ship Types-Alternate Homeport Analysis-Add Granularity Detailed Routes, Detailed Cost Variables
-Model Risks of traversing Bermuda Triangle
References
-crusing.org/regulatory/issues-facts-cruiseweb.com/royal-caribbean/eastern-caribbean/#itineraries
-vacationstogo.com/cruise_port/Caribbean.cfm-www.timeanddate.com/-www.latlong.net/ -www.cruisemarketwatch.com
-Professor Ned Dimitrov’s Brain
Questions?Cruise Liner Schedule
P