how can you predict when you land at airport XX although you've changed flights 5 times?
>scientifically discuss logistics
>scientifically discuss
>scientifically

honestly there are just factories of people that sort shit for trucks 3-4 times a day, and then a preload in most places, some nodes along the way are automated, some are post air mail, the point is each mode of transportation is a puzzle piece with a logical numeric value, and useing all of your pieces you
A. Have a computer plot out the best possible route( UPS implemented this for small brown trucks), and international node transfers alike.

B. You operate in a more domestic capacity, and have each mail truck play zone defense on the mailboxes.

>>7652344
Also theres a fucking ton of auto scans along the way so your logistics division can just mainline data livefeed, generate a plan of action based on the past, and future trends.

>>7652293
Le statistical data from le past.

You can expect the time to get from A to B at 2pm on a wednesday will have a narrow enough normal distribution to predict it with enough accuracy.


Statistics my m8

>>7652344
>honestly there are just factories of people that sort shit for trucks 3-4 times a day
Least we're not those fuckers over in ONT2.

>>7652360
Is it normal though? If it is late, won't it miss a flight and thus the events aren't independent?

>>7652366
if you have 15,000 flights from place A land at 10am land at place B at ~2pm, you say it will happen at 2pm even if 50 or 100 samples are much later e.g. between 3 and 10 pm

chances are they just discard the outliers, especially if it's for non-logistics reasons e.g. air traffic diverted or grounded for the day for X reason (presidential visit, terror attack, whatever)

>>7652293
newsvendor model
economic ordering quantity
poisson loss functions
normal distributions
knapsack problem
MC=MR

I assume that they have a huge list that gives every station a distinct number.

The algorithm compares the current station number with target sation number and if they are not the same it determines a random station where it is sent to.

>>7652870
So a random walk on a graph?