|
|
|
|
PHARE Demonstration 3 Experimental Design |
|
 |
|
|
|
The PHARE Demonstration 3 real-time simulations (plural site at CENA, EEC and NLR) included two organisations, baseline and advanced, and suitably structured training exercises.
PHARE Demonstration 3 baseline exercises related to the PHARE Demonstration 1 and PHARE Demonstration 2 reference organisations. This was a “strip-less” organisation which included some basic tools to assist the controller in planning traffic through the sector and also into the approach area.
Traffic mixes were designed to include 4D Experimental Flight Management System-equipped aircraft with traffic samples ranging from a 30% to 70% ratio of 4D Experimental Flight Management System-equipped aircraft.
|
|
|
|
The Advanced concepts to be evaluated included:
- human centred approach for automation;
- multi-sector planning environment;
- traffic organisation and 4D trajectory negotiation.
|
|
|
Airspace data for the reference organisation was taken from the CFMU database as of 21 June 1996.
The advanced organisation used the reference organisation data as start point but evolved in accordance with the PHARE Demonstration 3 operational concept to cater for future airspace structures.
The predicted annual traffic growth used in the PHARE Demonstration 3 demonstrations considered that by 2005 traffic would multiply by 1.5 on 1995 levels and that demand in 2020 would be multiplied again by 1.5. To cater for these traffic predictions the following series of traffic increments was defined: June 1996, 1.5x, 1.75x, 2x, 2.25x. The 2.25xJune 1996 traffic sample was not used with the Baseline system.
An initial experimental plan was defined on the basis of 14 Runs defined on the above growth hypotheses and taking into account the aircraft mix hypotheses. This plan was later further refined and adapted by the Validation Project.
|
|
|
CENA (Centre d'Études de la Navigation Aérienne) |
The PHARE Demonstration 3 main phase demonstration took place over six weeks in May and June 1998.
For system organisations were defined: the Baseline close to today’s operational with limited planning aids; the Advanced organisation with 0% Datalink equipped aircraft (A0) which examined the effect of the introduction of PHARE Advanced Tools and a new Ground HMI; and the Advanced organisations 30% and 70% DL (A30 and A70) which examined the effect of introducing aircraft equipped with 4-D Flight Management System and Data-Link. The objectives of the trial, namely to measure the impact of the PHARE Advanced Tools, 4-D Flight Management System and Datalink, were met by comparing the results from the various organisations whilst keeping other factors, e.g. controller rotation, system organisation and traffic volume, the same.
The traffic samples were generated from data corresponding to the 21 June 1996 which was the summer peak for French ACCs. The samples were generated by building up a reference from which the others were derived respecting the increasing ratio recommended for PHARE Demonstration 3 trials.
However, the proposed ratios could not be maintained without changing the space model for PHARE Demonstration 3. Therefore only the Departure flight numbers were increased which had consequences on the traffic load in the Extended TMA and En-route sectors.
Each controller participated in two weeks of measured runs and performed alternate role (either as Planning or Tactical Controller) on the same sector for each run.
Controller rotation has been chosen so that statistical tests could be run by reducing the effect of variability between controllers.
In each first week, Monday was used for refresher training, and during the other eight days a total of twenty simulation runs were carried out.
|
|
|
|
The design of the PHARE Demonstration 3 experiment at CENA is described in more detail in:
|
| |
 |
99-70-01 Volume 2 of PD/3 Final Report - CENA PD/3 Final Report Annex A Experimental Design and Methods69 Kb |
|
|
|
NLR (Nationaal Lucht- en Ruimtevaartlaboratorium) |
In the case of the PHARE Demonstration 3 project a series of main experiments was planned to take place in May 1998. These were planned to be preceded by two weeks of pilot trials in which the readiness of the simulation environment and the various people involved would be tested. The main trials were planned to consist of five consecutive periods of two weeks. In each period another international team of air traffic controllers would first go through one week of training and then one week of measured experiments.
With the problems arising from the development of the Ground HMI, the whole series of experiments at NLR was seriously affected. While there was still hope for delivery of the Ground HMI, the pilot trials and the first series of experiments were cancelled one after the other. When it became clear that the Ground HMI would not be available at all, it was decided to keep the last two-week period as a pilot demonstration of the baseline system and as a general rehearsal in case new experiments could be organised.
Directly after the end of the PHARE Demonstration 3 project the decision was made to develop a GHMI for the advanced system based on the PHARE Demonstration 1 Ground HMI which was already integrated in NARSIM.
New experiments were planned for early 1999. This would be one series of experiments lasting four weeks. The first week would be dedicated to training and the remaining weeks would be used to perform a series of experiments. Only one international team of controllers would participate.
While the development of the Ground HMI was performed according to schedule, it became apparent in October 1998 that there would be a funding problem for the experiments in 1999. The planned system integration trials in November 1998 were then changed to a final public demonstration. Over a period of two weeks a team of fourteen controllers from different states visited NLR. These controllers were trained on the operational concept and the use of the system. Then over a period of three days presentations were made to visitors from all over Europe.
|
|
|
EUROCONTROL Experimental Centre |
A planned schedule of exercises was defined for PHARE Demonstration 3 at the EUROCONTROL Experimental Centre (EEC). This aimed to minimise the effect of natural controller learning during the experiments on the objective results as much as possible by mixing the different experimental organisations throughout the simulation period. Within each airspace group, a small number of feed sectors was also defined and each controller was assigned to either a measured or feed position within their grouping for each simulation exercise.
The planned controller rotation sought to satisfy two objectives:
- allow the controllers to experience the different roles associated with their particular grouping, and thereby enhance the validity of verbal and questionnaire feedback
- respect the need for ‘balanced’ exercises in order to permit statistical analysis of the different simulated organisations
|
|
|
It was planned to use ‘observers’ as part of the data gathering strategy. It was intended that these observers monitor the working method of the controllers, noting any difficulty in the use of the system, and subsequently discuss problems with the controller upon completion of the exercise. Due to the fact that all of the exercises were conducted as ‘non-measured’, a more informal interaction between controller and observer was possible and a large part of the feedback came from these observations.
Approximately 32 exercises were run in total over four weeks, the first two weeks concentrating on the baseline and the second two weeks on the advanced system. However, no data was gathered during these runs for analysis.
|
|
|
|
|
