ATC Training: NATS Looks to the Future

NATS

Air Traffic Control

Today over 150 students per year from all over the world pass through the College’s door. Those destined for NATS operational units will be part of a team that handles a combined total of 6000 movements per day in complex and crowded airspace. In order to achieve a skill level commensurate with such a demanding control environment, today’s students must have access to the very best teaching aids - modern classroom facilities and teaching techniques, simulators that accurately represent the airspace and traffic they will experience in real life, and lessons and training exercises that are designed to fully prepare them for all eventualities they will experience in the “real world”.

However, today’s traffic levels are expected to double by 2020 and the European Commission’s introduction of the Single European Sky will greatly widen the horizons that the controllers must look towards. It is important that ATCO training constantly evolves and improves to prepare the controllers, of both the present and the future, for the challenges ahead – taking full advantage of technological changes to get the best from the best.

Air Traffic ControlTechnology

Over the past fifty years the technological advances within ATC have been tremendous and it is important that the training facilities keep pace with the change. Today’s students will often be taught on radar systems indistinguishable from those found in operational units, they may well have access to 360° Tower simulators, and they are able to hone their skills in a realistic representation of the real world. But what systems will be available in the next ten to twenty years, what will be driving their requirements?

There are two main drivers to the development of simulation technology within ATC training – the need to parallel the developments within operational ATC and the cost of acquiring and running the simulators. It has long been recognised that reducing controller workload is the key to increasing capacity while simultaneously reducing delays. Eurocontrol’s Programme for Harmonised ATC Research in Europe (PHARE) brought together the ATC research establishments of the UK, Netherlands, France, and Germany with their own at Brétigny to develop a prototype ATC system for implementation in the 2015 timeframe. The result was a paperless, stripless control process making extensive use of Computer Assistance Tools (CAT). Using CAT alerted the controller to possible problems and assisted him/her in their solution.

Today we are already seeing the influence of PHARE in the next generation of ATC systems. Within NATS, the elimination of paper flight progress strips and the introduction of CAT are evident in the Interim Future Area Control Support Tools (iFACTS) programme. Due for implementation at NATS London Area Control Centre (LACC) at Swanwick by 2008, iFACTS will provide the controllers with tools that will reduce controller workload and hence provide a capacity increase. The proposed system makes use of sophisticated trajectory prediction (TP) and medium term conflict detection (MTCD) to support a number of controller tools that will be displayed on the radar screen (see box).

Air Traffic Control

At an appropriate time prior to the implementation of such tools, students passing through the training colleges will need to be instructed in their use and in the associated control techniques. This, however, raises an important issue – that of transition. How do we teach students on an advanced system that is going to be introduced in a few years time, when we know that to begin with they will have operate on today’s system? At what point in the training process do we swap from today’s strip-based system to tomorrow’s stripless system and how do we conduct the necessary On-the-Job Training?

The resource cost of acquiring new and up-to-date simulators is constantly escalating. These simulators tend to be resource intensive in terms of both manpower (i.e. pseudo-pilots and instructors) and space - a 360° Tower Simulator tends to employ a larger number of cubic metres per student trained when compared to, say, an approach radar simulator. Needless to say, these increases in resource costs translate into financial increases at a time when the capital costs of the simulators may well be decreasing. It is therefore important to concentrate on reducing the resource related costs.

Latest advances in the field of Automatic Speech Recognition (ASR) have made possible its expansion into applications such as voice control, speech to text conversion and speaker verification. Discussions within NATS have identified the following areas for potential recognition applications within the ATM environment:

  • Automated pseudo pilot data entry for simulation
  • Time stamping of instructions for aircraft departure standard information
  • Datalink voice interface
  • Recording of aircraft nominal intention from ATCO instruction
  • Read-back error alerting
  • Callsign confusion alerting
  • HMI activation

It is clear that the proposed Direct Voice Input (DVI) applications could be advantageous to NATS in terms of cost-saving, commercial and safety benefits. Nevertheless, some of the identified applications are believed to be beyond the realm of current speech recognition capability. A research programme has been initiated that will support the development of DVI technology to maturity suitable for ATM applications. It will take a phased approach, initially targeting low complexity, non-safety critical applications within the training field. The experience gained will be used to address more ambitious and progressively safety-critical applications without disproportionate technical or financial risks.

The application of DVI in the training field will look at automating the task currently performed by pseudo pilots. Currently, pseudo pilots manually enter data to drive simulators during training or trials, so that simulated aircraft respond to ATC instructions. DVI of these instructions, possibly together with COTS speech synthesis, could lead to cost-effective automated simulators. The first phase in the programme is to introduce DVI on a Part Task Computer Based Trainer, used by NATS’ ATC students, to create an automated pseudo pilot – this will provide feedback to improve the DVI algorithms. The project will then move on to apply and validate the use of a DVI-activated automated pseudo pilot on low criticality applications, such as system testing of simulators and trainers, through to its application within benign (non-operational) ATC Environments such as research simulations and the complete range of simulation work associated with student training at NATS College of ATC.

But the use of DVI as an automated pseudo pilot must be only the first step. At NATS College of ATC, all simulation training is conducted on a one-on-one basis, i.e. one instructor for every student manning a simulation position. This ensures the best quality training possible – with the instructor able to give their full attention to an individual student. As instructors are frequently qualified ATCOs and could be used in an operational environment, this is a significant drain on a resource pool that is likely to be experiencing a worldwide shortage for the foreseeable future. Can we, therefore, envisage intelligent training simulators that not only react to the student’s voice and respond to his/her commands, but also observe the student’s performance and offer advice and criticism? At what point will artificial intelligence reach the level where computers can assist and work alongside the human ATCO instructor?

Space can be a pressing problem in ATC colleges – and many of the newer simulators are demanding more, not less, room. The use of Virtual Reality (VR) has often been touted as the next step in aerodrome simulation development, but it has proved to be a lot more difficult to implement than first envisaged. Using VR in ATC training presents a lot of problems; to train for the operational environment the student has to learn to interface with physical systems - telephone, information displays, flight progress strips, etc. - whilst also monitoring the situation outside the tower. At the same time, the instructor must be able to see what the student is observing and to point out what he/she may be missing. Such problems present unique challenges for VR developers.

As part of their Long Term Research programme, NATS Department of Analysis and Research has been looking at the use of Virtual Reality in the training simulation environment. A feasibility study was conducted into using VR Technology with a head mounted display to develop an aerodrome controller-training simulator. The main objective was to gain a better understanding of the potential benefits of this technology and to make well-founded recommendations for the future development of aerodrome simulators. A number of features were built into the demonstrator, including:

Air Traffic Control

  • A high-quality 360 degree photo-textured environment of the ATC tower and surrounding aerodrome (including buildings, runways and taxiways) as seen from the real tower;
  • Moving vehicles – including aircraft carrying out approach and landing, takeoff and taxi manoeuvres and ground vehicles moving about the airport surface;
  • Various weather scenarios including rain, fog and night/day emulation;
  • A see-through capability in the headset to enable the trainee to see desktop instruments and flight strips when required;
  • A trainee monitoring option, with the instructor having access to the views of the trainee as well as a global view of the airport.

This feasibility study concluded that the system has development potential and further work has since been undertaken to look at different options for the display interface within the same space constraints. The next steps would be to integrate the demonstrator with NATS traffic generator to verify interoperability and performance and to quantify the training benefits offered by this technology. The question of how to integrate such a tool into the overall training programme will require careful consideration.

Environment

Technology is not the only factor changing the face of ATC provision, and hence ATC training, the political and regulatory environment is exerting a significant influence. The advent of the Single European Sky will inevitably change the dimensions and nature of airspace boundaries, and consequently the interfaces between controllers working within that airspace. The current, seemingly inexorable, proliferation of smaller and smaller sectors in pursuit of additional capacity will eventually give way to fewer, larger sectors, aligned to traffic flows rather than national airspace boundaries. Even the “virtual” ATC Centre, staffed by controllers at different locations throughout Europe, becomes a possibility. With these changes will come changes to the controller’s skill and knowledge requirements. Just as the controllers we trained in 1949 were not equipped to cope with the ATC of today - extensive airway systems, small sectors, multiple co-ordination - so the controllers we train today will, unless we do something different, lack the wherewithal to cope with the ATC of tomorrow.

The harmonised European ATC Licence, scheduled for introduction at the end of 2003, is the first step towards a pan-European controller workforce. Along with the harmonised ATC Licence comes a need for harmonised ATC training, and the requirement to train to a common standard - Common Core Content – has already had an impact on ATC training providers across Europe. However, this is only the start of a sea change in ATC training, inevitable and essential if we are to prepare controllers for a very different ATC environment. It is too early to determine just what these changes may be and quite what training programmes may look like in a few years time, but NATS Human Factors Department recently completed, and has now implemented, two projects to help shape those changes.

Based on the premise that to effectively train new controllers it is first necessary to understand: what a controller does; how he/she does it; and why he/she does it – NATS Human Factors Department undertook an internal project to provide the fundamental building blocks of ATC learning. These building blocks only not encompass the knowledge and observable skills, but also the cognitive skills and attitudes required for a good controller. This project resulted in the development of a database outlining the building blocks required for ATC training. A gap analysis then identified where we are and where we want to be along this training roadmap – thus determining the new knowledge, skills and attitudes set, along with the training required to fill the gap. This work was then progressed through development of the Eurocontrol SHAPE framework, which assessed future ATC related technologies and the controller skills required to implement and operate those technologies. This has enabled the setting of targets to identify the skill change required in future controllers and how their training will have to be modified to develop those skills.

Conclusion

New technology, new airspace, new regulations all contrive together to produce an ever-increasing level of complexity for the controllers to cope with. There will shortly come a time – indeed many believe it has already arrived – when controllers will have to specialise, not just when they go operational but from day one of their training. The current controller training tends to follow a well-worn path of theory leading to aerodrome training, approach training (procedural followed by radar) and then finally onto area training. In today’s complex airspace, the area controller rarely, if ever, draws on his aerodrome skills, so why are these still so often taught?

The world of ATC is rapidly changing – new systems are coming on line and a revolution in operational methods is on the horizon. Training programmes, methods and techniques must not just keep pace but precede developments in the field, and innovation and vision in ATC training is just as important as innovation and vision in ATC provision.

This is an update of an article originally written by the author in November 2002 for publication in Air Traffic Solutions.

page up
×