I was able to get in touch with a German authority which has access to the data needed. I received the data for all arriving flights to Frankfurt (EDDF) for the year 2019.

I tried to predict the parking positions based on the aircraft type, the aircraft wake turbulence category, whether it is a holiday in Germany or Hesse, the departure airport, whether the departing airport belongs to the Schengen area, the temperature, the visibility, and the wind conditions.

The H2O (www.h2o.ai) software with the Gradient Boosting Machine-model was used (https://docs.h2o.ai/h2o/latest-stable/h2o-docs/data-science/gbm.html).

However, even after estimating a few million models with cross validation, no model was able to provide adequate estimations.

Below you can find the Scoring history in tabular format for one of the better estimations:

1   0.101113
2   0.102423
3   0.103255
4   0.105761
5   0.107898
6   0.108273
7   0.110989
8   0.112574
9   0.118112
10  0.120291

 As you can see, the first guess of the model is correct in only 10.11% (first row). Even if the model would be allowed to guess ten times, the probability to guess the parking position for arriving flights correctly is still just slightly higher than 12% (last row).

Thus, there seems to be either a huge undetected influence (e.g. the next destination of an aircraft, which was not available from the data) or the position assignment is more or less random in the real world.

In conclusion, as long as no other high frequency, real world data source becomes available, there is no way to estimate parking positions for arriving flights precisely.

* However, as there is a “bug” in the code parsing the TELEX-messages, no Connex Infos are sent right now, anyway. (https://forums.flightsimlabs.com/index.php?/topic/31373-efb-and-connex/&do=findComment&comment=249614)

in the recent past I have tried out a few things regarding web development. 

As a small result, I transfered the Trim Calculator App from an Android App to an Web Application.

For those of you interested, it can be found under the following link:

http://6bf5de.online-server.cloud:8099/TrimCalc/Trim

Kind regards,

Jan