Modelling the gate process at a container terminal.

Introduction:
In this assignment you are asked to model the process of trucks arriving at a container terminal gate,
moving to a handling location where they can either receive or deliver a container from a stack crane,
and driving back to the exit gate to exit the terminal. The modelling of this system should be done in
Java, and the result to be delivered is the Java application and source code as well as a short report
describing the approach and answering the main research question.

Problem formulation:
Trucks arrive at the terminal in the pattern described in the attached file TruckActivity.xlsx. At the entry
gate, all trucks need to be processed in one of the gate lanes. At each lane, the processing time of a
truck in minutes is described by the random variable X1, where X1 ~ Gamma(9, 3), where 9 is the shape
parameter α and 3 is the rate parameter β, and E(X1)= α/ β.

The truck then proceeds to one of 10 stack modules. Driving from the gate to the stacks or vice versa
takes 3 minutes. The stack modules can be modelled as having one single queue with 10 servers each
handling one truck at a time. Any truck can be served by any of the servers. The service time X2 ~
Gamma(4, 2).

After service by the stack crane, the truck drives back to the exit gate. Handling time at the exit gate X 3 ~
Gamma(3, 3).

Assignment:
How many lanes at the entry and exit gate would you advise the terminal to use?

But based on what would you decide that?

Extra Information:

I give you an extra information you can use for the assignment to help you to answer this question: The
terminal operator has indicated that the number of gate lanes should be such that in the provided test
scenario, the total number of trucks queuing at the entry gate should not exceed 30 with 95%
probability, and the number of trucks queuing at the exit gate should not exceed 20 with 95%
probability. Within these restrictions the number of lanes should be as small as possible.

Now you have all the information to answer the above question.

So it is clear the processing time at gate in and gate out (Gamma distribution, tip: do not forget to us the
scale variant as described, for instance for entry gate Gamma(9,1/3)) and now you have the
requirements from terminal regarding the number of gates.
Can you help the terminal operators to define the optimal number of gates based on the above
requirements?

Mail :

Dear Madalina,
 
We don’t have any other input sets for the data, this file should work. If it has some problems for you,
you are free to modify it a little to make it work easier for you.
 
The 95% probability means that 95% of the time the queue should not be longer than 30 or 20. So you
need to measure how much time the queue exceeds 30 and divide that time by the total time of your
simulation. This number should be below 5%. When you execute multiple simulations on the same data
you can simply summarize the time and make the same calculation.
 
Let me know if it is still unclear or you have other questions.
 
 
BR,
Pascal