EuroHarmony Community Forums
Archive => EuroHarmony VA => Old Forum => Technical discussions => Topic started by: EHM-0188 Andy on March 08, 2004, 04:47:04 pm
-
How did/do you all fill in the number of passengers on your PiReps ?
Do you fill in a random number or do you use a tool that calculates it ?
-
I think the way it used to work was if you used the online pirep form then you can put whatever you want, and if you didn't put anything then it was automatically set as the aircraft being 3 quarters full.
If you used the flight logger to submit the flight then when you select the aircraft you used, the number of passengers was automatically set to an amount somewhere between 2 thirds full and completely full.
Thats roughly how it happened anyway, the numbers probably aren't exact.
-
I used a randomize formula in MSExcel to calculate the number.
I just wanted to know if anyone used a certain tool.
Or is there someone that can make such a tool ?
-
As Joe said, if you don't fill in anything in the PIREP form then it takes a random number. It takes around 80% of the aircraft's capacity if you leave the field empty.
-
Using a randomisation tool would not be very realistic unless you put a minimum cap on it......which i assume you did......for me i always fly with full load of passengers! :]
-
Originally posted by Anvaco
Or is there someone that can make such a tool ?
I will make one, it's pretty simple, and first version will be very basic. Just a minimum and maximum number of pax, and generated random nr. in that interval.
Maarten, may I use EHM logo for this one?
To make it official maybe...
Cheers
-
And another quality product from EuroHarmony is in the works! ;)
-
Errr... didn't quite understand what this
randomization tool would do differently
from what the PiRep does automatically
already.
Marteen said that if the field is left empty
the PiRep automatically put something
-around- 80%. Which probably means a
random number between 70% and 90%.
Sooo... what should this new tool do?
Manu
-
Well......i guess it could allow the pilots to adjust the range according to current real world traffic trends? And of course it lets them know the number of passengers before hand and not after flying the flight! a tad bit more realistic i think
-
Manu, as I said it won't be anything fancy like logger was, nor will it connect through FSUIPC or something... Just a tool to create pax in given interval, though I am open to more suggestions that may be added to make it more complex.
Feel free to post them here.
And since I don't fly anymore, at least not nearly so often, I plan to contribute through other means, even if I don't have a diamond on my badge.
Cheers
-
I know of 2 existing tools for calculating the number of passengers.
1. Seats 2.0 from Michael Lyons
I tried it but I'm not able of changing the planes in the setup
2. Booking system from Ray Porteous for use with the VA Administrator software
(not freeware)
-
The idea of changing the number of
passengers according to the trends
is quite cool and wouldn't have to be
implemented on the pilot's side, it
could be done on the server side.
The problem is: how do you find the
current trends? How do you find how
many passengers tend to fly between
London and Paris at any given time?
Or for any other location served by
EuroHarmony?
Manu
Manu
-
That's a good angle, making 'trends', hmmm..
Maybe to make it simpler, we should divide destinations into classes also, say 6-10 different, and calculate approx trend, given departure and destination index and some constant.
Example :
Schiphol >Class 1 > coefficient 9
Okecie > Class 5 > coefficient 5
Plane class(capacity) approx. 220
Passengers leaving: 9*5=45 percent, (there could be interval also, +-3-5%)
meaning plane will leave with 99 pax for Warsaw .....
Just a basic thought.
-
That would be pretty elegant, but
if we knew so precisely how many
passengers fly between two location
just because of the size of the location,
we'd probably used a smaller plane,
wouldn't we? I wish we could find
some "live" data somewhere... or
past data we can extrapolate random
but not completely groundeless numbers
from....
Ciao ciao!
Manu
-
Hmm....or how about human labour? Form a group of people who decide how the current trends are running? Doesn't exactly fit into EuroHarmony's automation but well it could be a temporary solution till there is a live source found......
-
The problem with trends is that other than a major event at a route such as 9/11, trends tend to be airline specific. The trends for say EasyJet flying short continental routes from Stansted will be different from BA flight flying similar routes. Totally different marketting.
-
Then the trends for BA would apply to the Main Division while the trends for easyjet would apply to EuroHolidays or something! ;)
-
You could takea few (4-5) airlines for each division to calculate the Trend
Euroharmony ->
Average Trend of BA, KLM, AF, LH
Euroholidays ->
Average Trend of Thomas Cook, Virgin Express, Hapag-Lloyd, ...
-
And maybe also route trends or region trends?
-
Sounds awefully complicated maths for hat would in essence still be an estimation and probably not a whole lot more accurate than the calculation already used. After all if most seats aren't sold the route gets dropped.
-
True but in essence it would give rise to much more variations between routes, and i don't think EuroHarmony would play the game of dropping routes....
-
Originally posted by Plymyphil
Sounds awefully complicated maths for hat would in essence still be an estimation and probably not a whole lot more accurate than the calculation already used.
You are probably right ...
I made one just out of code, and it works, but I think most flyers will use it few times and then forget about it.
And since the logger sends data to server, there will be no use of it whatsoever.
Project halved, considering other options...
Hmmm, who's doing new EHM vTour?
I guess you have a lot more crucial stuff to think about, but I know I was first attracted here by it. If there are some text files left of the old version, I could make a new one.
Cheers
-
What do you mean with vTour?
Manu
-
Don't see any new tours around..........
-
what about this....
At 00:00 GMT, a certain amount of passengers are placed on each route... for example, 100 people. When someone chooses to fly on this route, the pax will automatically be set to X. This X is calculated by taking the 100 people (or whatever max is of the rute) and dividing it whith the previous 24 hour flights on the route. Example: Today five flights are made on a specific route. X=100/5=20. Tomorrow a flight on that route would be loaded with 20 passengers. I'm kinda tired now but wouldn't that after two days turn into a dynamic system, that changes day by day? Few flights previous day = fewer passengers = cut-downs on flights (fictional) = more passengers on the lasting flights = new flights to cope with the passengers (fictional). Wouldn't it work? Like in real life! :)
thank you.
-
Thank you fimfim, and welcome aboard.
A dynamic system like the one you propose
it's kind of interesting, because it would free
us from the need to collect a lot of real world
data.
On the other hand, any dynamic system will
have balancing problems. Think about having
a bunch of bottles half-filled with liquids
and connected by tubes to one another.
Then think about bringing the whole thing
on a ship in high seas. The liquids will flow
caothically from bottle to bottle without
too much logic, up to the point where big
bottles might be completely empty and
smaller bottles might be overflowing.
Out of the metaphore, we might end up
with big cities with no demand/supply and
small cities with an insane amount of
passengers flowing in and out!
Anyway, it's a very interesting and different
idea from the direction we were looking at
in the previous posts.
Maybe some kind of hybrid solution will
do. Something to think about.
Thanks for your input.
Ciao ciao!
Manu
-
I had been thinking of something like that too but because of the problem i did not mention it........besides that could call for limiting the routes that can be flown to the pilots which is not very good........the most realistic solution that i can see would be to find out the current trend, which gives you the demand for that route, and then divided by the number of flights flown that day.......which creates a hybrid of the two systems, but essentially combines the problems facing both solutions........back to thinking again!
And oh about fimfim's suggestion again, what if the next day only 1 flight was made, doesn't that mean that all flights on the next day would be loaded with 100 passengers? and if like 7 flights were made on the third day, wouldn't that be 700 passengers carried total, far exceeding the demand? Therefore i think that the number of passengers onboard should not be dependent on the previous day's flights........
Instead how about leaving the passengers carried blank and calculate it based on the number of flights that day and fill in the value at the end of the day? But then in this way the pilot won't know how many passengers they've carried till the next day.......
-
Wassup guys... I have been following ur conversation...
I'll make this very simple....
On the euroharmony website, there is this link called "PILOTS".... ok..
When you click on that, magicly appears the link "FLEET" ok...
by some weird circumstances, on the FLEET pages, all the EH planes in use are listed...even weirder, is that there a page for every on of this planes.... ok..
Now comes the good part.....
On each such pages, there a complete list of the planes specs. ...ok ..
Now comes the realy good part...
In that same specs list, theres that little section that says : Capacity (PAX) ...
And there it is people... the max amount each plane can have... so if it says 25, you can put 25 or less....it's ur own decision....
want an example..??
BOEING 747-400:
Planes In Fleet 16
Length 70.67 m
Wingspan 64.44 m
Height 19.41 m
Engines RR RB-211
Power 4x 266.9 kN
Cruise Speed 490 kts
Service Ceiling 45000 ft
Range 7135 nm
MTOW 385.5 t
Capacity (pax) 416
I hope I've Helped you guys out a bit ..
Have Fun Flying :]
-
uh.........i think we all know that! ;)
We're just discussing ways to generate passenger numbers on board a plane in a dynamic and realistic way
Was just wondering, what's weird about displaying all the aircrafts on that page with a page for each and everyone of them?
-
Hello,
I think it would be too hard or even impossible to get real life frequently updated statistics of every route. Of course, if it's possible then it's the best. I think we are limited to the EuroHarmony world because of the statistics we get from the pireps.
Or - why not get real statistics for 2 weeks only plus certain holidays, and then loop it? So that the PAX would be determined from holiday or not/day of the week/time of the day and such. To that som deviation could be added so the numbers wouldn't repeat completely.
That wouldn't be 100% realistic as the traffic on a route may change, but it could be updated every 3 month or something , that update would also help summer/winter-differences (charter routes and such).
Thank you.
-
Originally posted by manu3d
What do you mean with vTour?
Manu
The Virtual one, I thought it was lost.. good that it isn't ;)
I was away for the week-end, sorry for late replies...
See new ideas about the topic here.
Could some min and max amount of pax be set, to stop the over(under)flow, maybe?
Cheers
-
Fimfim, do you know where those two weeks
worth of statistics can be found?
Dejan, can you please rewrite your last
sentence in a slightly more understandable
and detailed way? Pleeeeeeease? ;)
Ciao ciao!
Manu
-
Originally posted by manu3d
Dejan, can you please rewrite your last
sentence in a slightly more understandable
and detailed way? Pleeeeeeease? ;)
Manu, Dejan's final sentence makes perfect sense to me (but then, I am a little twisted...)
I think what Dejan is getting at is (and correct me if I'm wrong Dejan): Would it not be possible to set some minimum passenger values for the aircraft in the fleet (maximums are already listed, Dejan) and that way a weighted bell-curve could be applied (airlines prefer to have too many passengers for a flight...) to get a figure for souls on board.
-
or how about this, take the trends that is in the continent itself (continential trends) and insert a multiplier for town to towns, city to city, city to town etc and derive the percentage of seats filled? and of course for long hual flights have intercontinential trends rather than route trends, but route trends can be thrown in for more prominent routes too........and of course festive seasons could be factored in as well but that could be covered in the continential trends already.......and of course last of all a random generator that generates numbers close to that derived from the previous workings...........think this would provide for the least work but yet the most realistic approximation
Anyone has a better idea?
-
Originally posted by manu3d
Dejan, can you please rewrite your last
sentence in a slightly more understandable
and detailed way? Pleeeeeeease? ;)
Sorry for not being descriptive enough...
Originally posted by Murray CraneI think what Dejan is getting at is (and correct me if I'm wrong Dejan): Would it not be possible to set some minimum passenger values for the aircraft in the fleet (maximums are already listed, Dejan) and that way a weighted bell-curve could be applied (airlines prefer to have too many passengers for a flight...) to get a figure for souls on board.
Yep, that's what I meant...
Just a thought to divert from idea of taking the real world data.
Correct me if wrong..., but I think depending on those data, means constant update chasing, and it will require a lot more work and effort.
:|
-
This is all sounding scarily like what (I believe) AirSource is trying to get at...
-
AirSource?
-
Originally posted by EHM-0962
AirSource?
http://www.air-source.us/
They brand themselves as "... a team of dedicated Virtual Pilots spanning the globe. Over 400 Airlines have contracts with Airsource. Every day, there are Pilots that call in sick, take vacation, go on strike, or simply cannot make it to the gate on time for their flight. AirSource is here to keep scheduled Airline service uninterrupted."
-
? i don't get it.........how does it sound like our discussion here?
-
Originally posted by EHM-0962
? i don't get it.........how does it sound like our discussion here?
Their back-end software inserts passenger numbers for their website...
-
Fim fims idea of based on previous flights sounds great but with a randomised function could create very unrealistic totals. If the function returned a slightly lower number than average then the number being used for future flights would also be lower resulting in a downward trend for that route. Ultimatally we could end up with a large plane and almost no passengers which would be highly unrealistic and against the grain of what any routine like this is supposed to achieve.
What are the industry averages for %pax. These must be available somewhere and if not could be obtained just by taking a poll of routes we have fly in real life and the number of empty seats. I would guess that the figure rarely drops much below about 80% and is probably usually between 95 and 100%. It would certainly follow a normal distribution curve so you would find say 95% of the flights are within the 95-100% capacity pax band. (These are my guesstimates and need testing.) In which case the solution could simply be a controlled band for a random generated number. 95% of the time then, the pax would generate in the band indicated and the rest of the time outside the band. The maths is fairly simple it's just finding the data to produce that normal distribution curve that might be more tricky.
-
Originally posted by Plymyphil
...I would guess that the figure rarely drops much below about 80% and is probably usually between 95 and 100%. It would certainly follow a normal distribution curve so you would find say 95% of the flights are within the 95-100% capacity pax band. (These are my guesstimates and need testing.) ...
For my own flying, I think I'll probably just make a simle program, giving me a random number within the range you described, I personally don't thing that the pax is so important to be exact numbers following the real life trend. It just seems too much work and too little to gain. But again, if it's possible to do it in an easy and elegant way, why not. :)
Thank you.
-
Originally posted by fimfim
For my own flying, I think I'll probably just make a simle program, giving me a random number within the range you described, I personally don't thing that the pax is so important to be exact numbers following the real life trend. It just seems too much work and too little to gain. But again, if it's possible to do it in an easy and elegant way, why not. :)
Thank you.
In which case fimfim why not just let the PIREP do it for you. That's more or less what happens if you leave the pax blank on the PIREP. It does a calculation of I think between 75%-100% and fills in the blank.;)
-
I've done it so far. :) But that way I don't know anything about my dear passengers, I let them in, fly from A to B, and let them out. But after all, that's what air travel is all about..hm.. Maybe cargo is more my thing.. :)
-
Originally posted by fimfim
I've done it so far. :) But that way I don't know anything about my dear passengers, I let them in, fly from A to B, and let them out. But after all, that's what air travel is all about..hm.. Maybe cargo is more my thing.. :)
We need some sort of conversation parsing software to generate chat as they board and maybe it could also create tickets with simulated addresses. Hey, one of us could check the simulated passports match the faces.;)
Better still, we could have a simulated drunk trying to board and then we could in a simulated way of course, kick his butt right off our aircraft and refuse to fly him ... oh ... the possibilities are endless.;D
-
...and a virtual little kiddie wanting to take a photo of the flightdeck every two seconds :) "what does this button do?"
(nothing wrong with taking pictures of flight decks)
-
yeah i have to agree......if it's too complicated to make, then we better not make it.......
And oh for the average pax for real airlines, i believe they stand at about 75%......even an 80% would be considered to be very good unless of course it's a special occasion when many people wants to travel there........but then again we could simulate EuroHarmony to be much more successful that the rest! :P
-
Someone or I could just make a javascript file and within it have all the aircraft along with a max. and min. number for each aircraft. Then when you press a button with the airrcaft's name on it, it generates a random number.
-
Isnt that a feature already in the online PIREP form? ;)