that replicate the complex behavior of passengers, but a tools such as LEGION can help to get a good approximation for the
purposes of this research.
The LEGION software has been calibrated and validated for some scenarios (Berrou et al., 2007). In the case of Santiago de
Chile this process was done by Metro de Santiago. In our study, each passenger is represented in a two dimensional view
(2D) and the simulation runs in a continuous space in 0.6 s time steps. As an input data we used the Gaussian distribution
function for the assignment of size and speed of passengers. In this work output variables are measured in their maximum
values.
In addition, in this stage we mocked up in the HDL of Universidad de los Andes a hall of a metro car and the portion of the
platform with the same characteristics than the LEGION model. As shown in Fig. 3, the hall is 3.00 m long, 2.50 m wide and
2.40 m high. The platform was 4.40 m long and 4.00 m wide. The 4.00 m width was chosen because it was the optimal width
found in Fernandez et al. (2010b).