37TH ANNUAL CONFERENCE, Toulouse, France, 30 March – 3 April 1998
WP No. 98
Short Term Conflict Alert
SC 1 was asked by PCX / CEO of IFATCA to look into the subject of STCA and develop policy on the matter when necessary. Also to study, based on new STCA policy, the relation between STCA and TCAS.
IFATCA has extensive policy on TCAS. On the subject of STCA, or other ATC conflict detection systems, no IFATCA policy exist.
Regarding TCAS, world-wide ICAO SARPs are established, referring to the same TCAS. As far as STCA is concerned, no world-wide ICAO standards are available and there is no uniqueness in the use of terminology for STCA.
In the line of defence to prevent collisions, STCA and TCAS are safety nets that are not yet considered as a safety-critical part of the ATC operations. The ATC environment can have different layers of defence: ATC procedures, flow control regulations, Plan Conflict Detection systems, Medium Term Conflict Detection systems and, as a final layer in the ATC system line of defence and the last opportunity to warn the controller of a possible hazard, Short Term Conflict Alert (STCA) systems.
Independently from the ATC system there is TCAS, an airborne system and a last resort and safety net for pilots to prevent collisions.
STCA is a ground based system and provides controllers an alert on the radar screen of future separation infringements between two aircraft, using track data supplied by the radar processing system. It is a computer program that is designed to look ahead in time (± 2 minutes) and detect potential conflicts between two or more SSR radar blips by comparing and tracking the horizontal and vertical evolution of all aircraft that are meeting the criteria for STCA detection and alerts. These criteria can be set as parameters, unique for every ATC facility.
Until some years ago the detection of conflicts between two or more controlled flights was done in a manual (‘human’) process, with no support of any automated detection systems. There was no back-up or safety net available to provide controllers with a final warning when they failed to detect conflicts in time or in the event of errors of judgement or system failures.
With the constant increase of air traffic, world wide air traffic has doubled in the last 10 years, enormous strain was put on the ATC system to increase the capacity of control sectors and on the controller to increase his productivity, in the sense of handling more traffic. The call for back-up and new safety tools in ATC systems became stronger and stronger. With the introduction of ATC systems that provide flight plan processing and radar data processing (tracking) and the association of a radar track with the appropriate flight plan (and automatic flight plan updating) the workload of controllers was reduced, but that reduction was nullified by the ever increasing traffic demand. Programs were developed to provide controllers with a safety net: a computer program that presents to the controller a possible conflict; it works in the “background” without interfering with the controller’s normal work, and without demanding manual inputs. After a period those programs were referred to as Short Term Conflict Alert.
As there is no common unique terminology for STCA, the systems can be designed with different operational objectives. It can be designed primarily as a safety net against collisions or as a back up for separation assurance and/or quality assurance.
If STCA is designed as a separation assurance tool for the controller, it should be able to predict potential violations of the separations standards. The look ahead time has to be long enough for the controller to recognise, evaluate and decide on an appropriate instruction (clearance) to prevent separation violation. In this design a lot of nuisance alerts will occur. And even controllers might use it as a “separation tool” (e.g. vectoring aircraft parallel, and when the alert goes, correcting the vector).
If STCA is designed as a safety net for the controller, it is oriented towards predicting aircraft in close horizontal and vertical proximity (a specified distance, closer than the prescribed separation standards). It can use rather short look ahead times, just sufficient for the issue of an instruction and execution of an avoiding action or escape manoeuvre, most probably resulting in a situation where two aircraft are manoeuvring in close proximity, a specified distance within separation standards but taking account of a critical miss probability, based on ICAO ‘s Target Level of Safety (TLS) criterion.
STCA algorithms can also be used as a quality assurance, whereby the actual minimum achieved separation between aircraft is continually monitored by computer programs and if the actual separation is below the established minima, it will be logged and/or an alert is given at a supervisors desk. For example the SMF (Separation Monitoring Function) in use in the UK since 1993. In the USA they have for more than 20 years this function built into their NAS and ARTS III. The controllers call it the “Snitcher”. In France they study all cases where the separation at the closest point of encounter was half the prescribed separation standard or less.
About legal aspects in the case of an alert of a separation infringement, which can result in an investigation, the question can be raised if the existing IFATCA C policy, (IFATCA manual 18.104.22.168 , 22.214.171.124., 4.4.A3, 4.4.A4, 4.4.A5), covers these kind of situations in their accident and incident investigation policies.
Longer look ahead times (dependent of the separation figure, that is put in the system as a parameter: the greater the separation figure, the greater the look ahead times) will result in more nuisance alerts to the controller. It is essential to ensure the optimum balance between providing genuine alerts with sufficient warning time versus minimising nuisance alerts. The warning time should be sufficient for the controller to comprehend the problem, to decide on an appropriate instruction to prevent collision risk, to communicate with the pilot, for the pilot the decision making process, and the time the aircraft needs to execute the manoeuvre.
The number of nuisance alerts is further influenced by the quality of the radar tracking system.
The alert message should be presented in such a way that it readily attracts the attention of the controller and that the problem is readily comprehensible. The STCA interface must be designed to support clear understanding by the controller of the problem and to enable the controller to control the situation positively without detriment to other concurrent tasks.
Alerts can be given visual (special attention symbols, blinking, violation graphics, velocity vectors, special colours etc.) or aural (audible alarms).
Other features STCA systems can have:
- selection of STCA mode
- acknowledgement of a STCA alert
- suppression of STCA alerts for a radar blip
- suppression of STCA areas
- STCA log
- supervisor output
It is important to involve controller representatives during the design and development phase and to provide proper introduction and training when implementing STCA systems.
It is important that for each individual ATC radar unit, parameters and nuisance filters in the STCA systems are being developed and tested, that they are suitable for the area involved, adjusted to the procedures, airspace layout, separation standards, radar source, traffic mix, etc. The systems logic and parameters should be flexible.
The relation between STCA and TCAS
In the ideal situation TCAS and STCA should be working independently and autonomous. STCA should give alerts to the controller before TCAS gives alerts to pilots.
However a UK CAA study (on behalf of SICASP) states:
“With TCAS Version 6.04A it is expected that STCA will alert the controller to a conflict sufficiently in advance to enable avoiding action to be issued which pre-empts a TCAS RA on approximately 55 % of occasions; on 17 % of occasions the STCA alert will occur ahead of TCAS but in insufficient time to enable avoiding action pre-empting an RA; whilst on about 10 % of occasions TCAS generate an RA at approximately the same time as an STCA alert. On the remaining 18 % of occasions there will not be an STCA alert. [This is due to the differences between the STCA and TCAS II systems, for example the TCAS interrogation rate is up to once per second and, as a result track profile changes can be recognised more quickly than by STCA which is dependent upon ground radar update rates around once per six seconds]
These figures are only an indication, it should be remembered that neither STCA nor TCAS II systems are aware of aircraft intent”.
The figures above indicate, even if they are only an indication, that STCA can result in a reaction and instruction by the controller that can be in conflict with a TCAS warning (RA) to the pilot. Also is TCAS generating RA’s to the pilot, without, or before any STCA warning to the controller.
In order to reduce conflicting warnings between STCA and TCAS, the systems should be designed so they can function independently and autonomously. A ‘natural’, time based sequence of the different warning methods should be, that after STCA, the last safety net of the ATC system, TCAS forms the last safety net for the pilot.
Alert times of STCA and TCAS are within limits variable and dependent on the parameters being used.
STCA logic parameters and variables are different for each area where the system is in use, depending on the sort airspace, radar sources, traffic density, traffic sort, procedures and separation standards.
TCAS logic parameters are dependent of the software version being used and the variables such as aircraft performance, indicated airspeed, climb/descend speed, closing speed, sensibility levels, distance modification.
Both systems do not have forward knowledge of aircraft intent.
Also sudden, unforeseen aircraft manoeuvres will not be triggered in time by TCAS or STCA to give alerts.
New technologies could make it possible for ground based “STCA-like” systems to use down linked aircraft data from the FMS; also for “TCAS-like-airborne systems” to use up linked data from ground based ATC systems.
Those “cross-bred” systems immediately generate, apart from technical questions, great problems regarding reliability, integrity, dependency, responsibilities of pilots and controllers and validation of the data being up/down-linked.
The up- and down-linking of data, especially the validation of the data, is a separate discussion subject, beyond the scope of this paper.
The usefulness of STCA systems has, despite the initial problems like the high incidence of nuisance alerts and false alerts (much the same problems as TCAS encountered at the beginning) never been questioned. The STCA, that could be called “the TCAS of ATC”, has seen in the last 10 years very significant improvements and it is believed that the installation of a STCA in each SSR equipped radar unit is nowadays a very significant step improving overall safety.
STCA programs are now considered to be mature enough to enhance overall safety in a modern ATC system by providing the controller a ground-based safety net. It is an indispensable tool that provides the final layer of safety that is needed in an ATC environment that is becoming more technology driven in more and more busier controlled airspace.
STCA systems should be considered as a last resort to advise controllers of potential losses of separation or collisions. It cannot be used to reduce separation standards or increase capacity (it is not a control tool) . The use of STCA programs as a quality assurance tool (not on controllers positions) should be studied by SC 4 and/or SC 7.
Last Update: September 28, 2020