NetLogo User Community Models
## WHAT IS IT?
This model simulates merchants vessels sailing from a west coast to an east coast port where there are pirates in the water. Please see the following paper for analysis conducted with this model: Sibley, C. (2016, June). Be Alert and Stay the Course: An Agent-Based Model Exploring Maritime Piracy Countermeasures. In International Conference on Social Computing, Behavioral-Cultural Modeling and Prediction and Behavior Representation in Modeling and Simulation (pp. 356-365). Springer, Cham.
## HOW IT WORKS
Merchant vessels have destination ports which they sail towards unless they see a pirate, in which case the merchant vessel will sail in the opposite direction of the pirate. Merchants have alertness values which are randomly assigned between 1 - 5 patches. This alertness value determines how far out the merchants will observe a pirate and begin modifying their route path.
Pirates establish a initial target merchant vessel which is the closest vessel to their start point. The pirate will pursue this target unless another merchant vessel comes in closer contact with them (i.e., is an easier target) in which case the pirate updates his target and goes after the closer merchant vessel. Pirates also have to beware of Naval vessels, since they can be arrested. Pirates can see Navy ships 5 patches out and will change their route, aka modify their behavior, by sailing in the opposite direction of the Navy ship to avoid being arrested. Only once the pirate is 5 patches outside of a Naval ship will they begin pursuing a new target.
Navy ships will stay staionary until a pirate comes within 4 patches of them or a merchant vessel nearby calls them for help, at which point the Navy ship will pursue that pirate unless anothe pirate comes in closer contact with them.
Once all merchants have either reached their destination port or been robbed, the simulation ends.
## HOW TO USE IT
The user determines the following initial numbers: the port count, merchant vessel count, pirate count, and Navy ship count, evasion heading. After setting these, the merchants are placed along the west coast and the Navy and pirate ships are randomly placed on any other patch.
The user also sets the mean speed of all three ships ( the merchant, pirate and Navy ship), which are randomly assigned values based on a normal distribution with user-defined mean and standard deviation of 0.1. . However, when the pirate or merchant are close to one another they both are hard-coded to speed up by 1.3 X their speed. The merchant speeds up when the pirate is within their alertness threshold and the pirate speeds up when they are within 5 patches.
The merchant-altertness-factor will multiply the alertness distance by the user-set value. For example, if a merchant typically can see pirates when they are 2 patches away, a merchant-alertness-factor of 1.5 will extend their vision to 3 patches and mean that the merchant begins mdifying their route behavior sooner.
## THINGS TO NOTICE
Play with the alertness-factor, which will increase the distance at which a merchant ship can identify there is a pirate nearby. You will notice that increasing the alertness actually increases both the percentage of merchants who are arrested AND the time it takes on average to get to port when there are few Naval Vessels, relative to pirates.
## THINGS TO TRY
See how different ratios of merchants:pirates and pirates:Navy impacts the percent of merchants who are robbed. Do more Naval ships mean fewer % arrested?
Also see how much speed makes a difference by adjusting those values.
## EXTENDING THE MODEL
Different geography will definitely impact the behavior of all the agents in this model. It would be iteresting to play around with different routes, for example, like navigating through the narrow passage ways and seeing where the best locations are to place Naval Ships to deter pirate activities.
Also adding in GIS is a natural extension for this model. Factors such as weather, bathymetry, and geography would certainly impact output.
## CREDITS AND REFERENCES
This model was inspired by a 2013 paper "Agent-based model of maritime traffic in piracy-affected waters" by Ondřej Vaněk , Michal Jakob, Ondřej Hrstka, Michal Pěchoucěk
(back to the NetLogo User Community Models)