29 Aralık 2007 Cumartesi


This is the third of my four articles serie on FALSE SENSE OF SAFETY
in the Air Traffic Control systems. The fourth article will be on
the "Passenger's Right to Know"...

I would like to draw your attention to a brilliant article on the situation awareness of the air traffic controller:

by Endsley and Rodgers (1996).

You can find this article on the internet.

Endsley and Rodgers article is about a scientific experiment done to measure the situation awareness of air traffic controllers. "A study was conducted to investigate the way in which controllers deploy their attention in processing information in en route air traffic control scenarios.(1)" The article is 4 pages long and explains clearly the terms and conditions the test was made.

The test procedure was described as: The subjects do normal controlling in an area for which they were licensed. "Each scenario consisted of a recreation of the ten minutes immediately prior to the occurrence of
the OE(operational errors). Twice during each scenario, the recreation was halted and the screen blanked."

After that:

"Subjects were asked to indicate the location of all known aircraft on the map, and, for each aircraft, to indicate or make a judgment of:

(1) if the aircraft was:
(a) in the displayed sector’s control,
(b) other aircraft in the sector not under sector control, or
(c) would be in the sector’s control in the next two minutes,
(2) aircraft call sign,
(3) aircraft altitude,
(4) aircraft groundspeed,
(5) aircraft heading,
(6) the next sector the aircraft would transition to, (7) whether the aircraft was climbing, descending or
(8) whether the aircraft was in a right turn, left turn or straight,
(9) which pairs of aircraft had lost or would lose separation if they stayed on their current
(assigned) courses,
(10) which aircraft would be leaving the sector in the next two minutes,
(11) which aircraft had received clearances that had not been completed, and, for those, whether the
aircraft received its clearance correctly and whether the aircraft was conforming to its clearance, and
(12) which aircraft were currently being impacted by weather or would be impacted in the next five

The article describes the results and evaluation also:
Subjects’ responses to each question were scored for accuracy based on computer data for each aircraft at the time of each freeze. Responses were scored as either correct or incorrect based on operationally determined tolerance intervals. Missing responses were scored as incorrect."

You can find the complete article at:

My point is: A similar experiment could be done for measuring SITUATION AWARENESS IN THE MAINTENANCE AND ENHANCEMENT OF ATC SYSTEMS. As an example, I will state below a number of SA criteria similar to the above given...

The persona who should be taking the test are:

1-Operationally responsible person from the software maintainer/developer
2-Operational interface person between the operational institution and the software maintainer/developer
3-their first aids or replacements
4-software configuration's responsible person
5-version control's responsible person
6-various group supervisors who are also quality inspectors such as Radar, FPP and ATN groups.

The questions they should answer could be:

1- How many releases are released in the last six months, one month, last week respectively?
2- How many errors and system abends were done for each month and each release in the last six months,
one month, last week?
3- How many operational deficiencies are solved in each release in the last six, one months and last week?
What are the operational and technical difficulties of these according to a well established complexity
criterium? (system LOC, change LOC, system complexity, changed region complexity).
4- What structure and content has the release? Multi-changes, small changes, big changes etc.
5- What direction is the system maintenance effort cruising in terms of four items above? More difficult, same or relaxed?

Human factors
1- What is the motivation level in the software team? Specially people who work on the current changes?
2- What is the number of hospital visits, doctor visits, nervous breakdowns in the team?
3- What is the number of sick days taken of by the software personnel?
4- What is the number of leaves taken from annual vacation?
5- What is the number of days that could not be used for vacation leaves although they were used?
6- What is the number of official complaints or telephone calls to the welfare services?
7- What is the total duration of time spent in technical and other meetings, including official but personal
8- What are the number and duration of telephone calls made in the team?
9- What is the daily and total working duration?
10- What is the number and duration of lates at lunch and other breaks?
11- What are the number of quarrels and heated discussions in the team?

Weather conditions
1- last week (sudden changes?)
2- at the planned time for the release.

Adjacent centers and other affiliates
1- Communication conditions
2- The availability of service at the adjacent centers 3- track record of adjacent centers' availabilities at that
time of the year

Military conditions
1- the level and quality of interaction with military
2- temporary reserved area situations
3- communication lines etc quality

1- system availability: hardware, planned maintenance etc.
2-personnel: leaves, new policies such as
personnel reduction, service time increase etc.

1- number of at that time of the year
2- track record of airmisses for the last
six months including the recent

1- Number of flights:general, sector, point
2- Flights at that time of the year and at that hour
3- Number of special events at that time of the year
highjack, emergency sicknesses etc.

Flights of airlines involved
1-Airlines related with scheduled flights at the release moment

Night - day conditions

Operational team - controllers' condition
1- strike
2- reduced team for some reason
3- workload and experience of the controller directly involved.
4- personal condition if any of the controller

As the saying goes "You can not manage if you can not measure". These questions and possibly others have to be detailed and distributed to the related members and supervisors of the ATC software-maintenance team. They should regularly answer these questions and an automatic system should produce a single number indicating the RISK of realising the changes at that moment of the time.

This number could be conveyed to adjacent centers so that they could calculate their own risks. All of this is possible and meaningful in relation to a series of values taken through the time(as in (1)).

These are simple principles to think but very "complex" things to realise. The reason behind this is ATC people are generally "too busy" and "overloaded". Namely, their motivation is low.

If anybody had known there were 3 levels of maintenance or tech problems going on in Switzerland, they would not let an airplane full of chieldren go there... People SHOULD HAVE the RIGHT TO KNOW. And we the engineers should provide the existing information with due transparency and respect to human life.

Kind regards.



1- (1996) In Proceedings of the 40th Annual Meeting of the Human Factors and Ergonomics Society ( pp. 82-85). Santa Monica, CA: Human Factors and Ergonomics Society.

Mica R. Endsley Mark D. Rodgers
Texas Tech University Federal Aviation Administration Lubbock, TX Washington, D.C.

2- Schneidewind, Methodology for Validating Software Metrics
IEEE Transactions on Software Engineering VOL 18 No 5 May 1992.