Design in research

Design plays a critical role in the adoption of new technologies. This makes it a key player in research.

The other 2D

The radar image is always horizontal 2D; the ASTER project offers a vertical view for visualizing radar control images during airport approach.

icône publication
Scientific publications

This project gave rise to many scientific publications and opened up issues which are still currently under investigation.

Overview of the ASTER device

Dispositif et interface du logiciel ASTER

A view of the radar control interface

Gros plan de l'interface ASTER

Note taking during manual input

Écriture au stylet sur l'interface ASTER

Handling electronic strips with a stylus

Manipulation d'un strip sur l'interface ASTER

Genesis of ASTER project

In 2004 Intactile DESIGN and Intuilab created, on behalf of Sofréavia (Egis), the touch interface for Maestro D-MAN - decision support software for air traffic control. In 2006, Intactile DESIGN was awarded a Star by l'Observeur du design for the interface quality of this product. Its success proves to people within ATC (Air Traffic Control) that design can help solve very technical workplace problems.

Intactile DESIGN is also associated with the research project ANIMS (2004-06). The results prove scientifically that animation and interaction design increase the readability of HMI and therefore, by reducing margins of error, the reliability of the tools.

The precursor of the Maestro D-MAN is the electronic strip management tool Digistrip, designed by JL Vinot and his team at ENAC. In 2000, they carried out the research project VERTIDIGI on a vertical radar visualization tool for approach control.

The ASTER project has emerged from these tools and offers the mix of vertical viewing and electronic stripping.

Why use designers as part of a research project?

Based on our successes, which demonstrated the value of design at the centre of research, we were invited to work with teams at ENAC-DGAC on the ASTER project. The Intactile DESIGN approach is clear: our aim is to do research "with" design, that is to say equip research with collaborative and creative methods specific to design.

In the aviation community, it is not standard practice to integrate design at the development stage of the interface: engineers design systems; ergonomists measure their qualities and faults. This gives rise to an acceptable interface from an ergonomic point of view and robust from an IT perspective... but not one that is necessarily easy to use, or satisfactory graphically. Indeed, any envisaged optimization is mostly technical. More subtle and potentially richer solutions through graphic design, interaction design, or animation are not necessarily imagined.

However, in such a sensitive area as air traffic control, it would appear necessary that the tool be thought out in every detail. As will be seen, it is not easy, there being a great variety of inputs and very many levels of information.

The line of research given us is to aid anticipation. More specifically, this is to identify and design "objects" that, in the context of collaborative work and non-verbal communication, would help controllers anticipate traffic overload or underload.

Context of use

Air traffic control has a strong professional culture, where every gesture takes on a measure of responsibility. Each flight has to be dealt with case by case. Decisions taken by the controller in front of his tool, ranging from the urgency of certain situations to regulatory safety margins, are crucial. He is responsible for guiding all pilots from one airspace to another. He has to detect recurring patterns and unpredictable situations. These are common and should be dealt with before becoming dangerous.

Tour de contrôle, écran de contrôle aérien et tableau de strip

Control Tower | Strip boards | Annotated strips

Each air traffic control workstation is shared by two operators: the planning controller, who organizes flights in the early stages and prepares the work of the second operator. The latter, the tactical controller, is in radio communication with the pilots. Currently the workstation has a printer placed between the two operators which regularly prints out flight information on paper bands called strips.

Each of these strips is a flight. They are indispensable to the work of a controller. Throughout the monitoring of a flight, they will be annotated, crossed out, rearranged, replaced, moved, handed from one person to another and thrown out. This wealth of input is basic know-how for any air traffic controller.

Terminal areas feed one or more major airports. They are the interface between "arriving" sectors and "approaching" sectors. Planes enter these sectors cruising at great speed and at a high altitude. Speed must be reduced and the aircraft brought to an agreed flight level. They will then be transferred to the approach sector when they fly over a specific beacon. Aircraft must at the same time be staggered as a delay between flights is applied to allow runway capacity to absorb traffic.

Currently controllers leave the radar image and move to the strip chart in order to build a mental picture of the situation.

Methods to provide coherent interdisciplinary work

The design work was based on an in-depth analysis of the controllers’ work, which highlighted the main weaknesses of the current system and paved the way for suggestions for improvement. In ASTER, a touch screen was positioned to replace the current paper strip board - the most natural place for a manual input tool.

Given the importance of the vertical plane in the work of separating and resolving conflicts between aircraft, the ASTER project started from the idea of increasing the vertical display of traffic information carried on the electronic strips. However, as things progressed work on the strips became more and more important – to the extent that we could mask the vertical view in the final version. Electronic strips are then handled by a stylus and controllers can enter clearances and instructions directly onto these digital objects.

Once these fundamentals elements were laid down, the sketching and storyboard, both flexible and efficient tools, helped in laying out the interactions quickly.

The relevance of each interaction was then evaluated at co-design meetings and promising solutions were given successive run-throughs until their development – the duration of which was controlled in order to obtain a satisfactory result fast.

Storyboard

Images originating from ASTER storyboards

When passing from paper to digital, the real challenge was to increase interactivity and speed of data processing, but WITHOUT losing the freedom of action and rapid assimilation when using paper - to sum up, ensuring that the digital tool did not weaken the efficiency of gestures and practices. Another challenge was being able to anticipate a situation - what clues could interfaces give to indicate, for example, up and coming traffic overload (or underload)?

Working closely together, side by side, implies gestural code and non-verbal communication. For example, a controller can follow the progress of a task on the other's screen, and could even realize that the other is making ​​a mistake. It is because of this closeness that gestures are more collaborative. In ASTER, it is the potential interactions with strips that bring about this collaborative dimension – note taking and input gestures but also transmission gestures like throwing an object from one screen to another, or even dropping a strip directly from one screen to another! Storyboards as shown above illustrate, for example, different transmission capabilities.

Historique des Strips ASTER

The information carried by the Aster interface is numerous and should be clearly prioritized relative to each user. So we worked on the graphical consistency of the whole system, taking into account the sector’s culture and graphic codes used in its work environment. As ASTER is a research project, successive design versions were evaluated through experiment - different design points were refined on issues identified during a one-to-one with potential users using conditions close to a real time situation (control room).

The story of the strip design, as shown on the side, allows one to measure the many adjustments that were made to successive versions. You can see how the graphics structure the information better, so allowing greater clarity.

Focus on three design issues

In the ASTER project, the question of design quality is integrated into the research process. It became an optimisation parameter in the same way as the strength of IT functionality or the accuracy of data display.

Strip states

From the moment a flight enters the control area and the moment it leaves its digital strip will go through many states.

Flight entering terminal sector

Flight transferred from the previous sector, ready to be supported

Flight supported by the Planner Controller (PC)

Flight selected by the PC

Flight annotated and transferred by the PC

Flight transmitted in the shared zone, PC side

Flight transmitted in the shared zone, Tactical Controller (TC) side

Flight currently unsupported by the TC

Flight selected by the TC

Flight assumed by the TC, by radio communication

Flight transferred to the next area (approach)

Strip readability must be immediate and their design must indicate their status. Among many solutions, it was suggested that each workstation be matched to a colour: blue for the planning controller and brown for the tactical controller. Two colours which mark the strips "belonging" to one side or the other – and for which the controller is responsible. As shown above, other ways are used to indicate different states – the materiality of the graphics, the level of information content detail, the brightness, etc.

The wealth of entries

Preserving the freedom to use paper strips was crucial to the project’s acceptance by the professionals – and was an interesting testing ground for project scientists and ergonomists.

Digital strips are, of course, real-time informational access points about each flight but they also serve as input points. With a direct gestural interaction on a strip, the controller can enter clearance information (orders to pilots) complementary to radio communications.

ASTER does even better. Just like its paper ancestor, a digital strip can accept handwriting. With a simple gesture, the controller can switch the interface to "free input" mode and write, draw, cross out or underline information on the strips themselves or on notes – which they can then attach to visible aircraft in the background of the vertical radar view.

L'écriture dans l'interface ASTER

Types of input and feedback: user data via gesture, system data in gray, hand writing on strips or notes

A precise typographic code identifies each type of writing and indicates the author (planning controller, tactical controller, system) and its kind of information (order, hypothesis, proposal, forecast). The design was finally contained and its efficiency optimized within each technological field - no anti-aliasing, performance issues, or number of colours, etc.

Adapting the design at the time of integration

Design nurtured by the work culture

Our first instinct as designers was to provide a work space on a light background, corresponding to the aesthetics of paper and which would make it stand out from computer technology. But the actual observation of an air traffic controller carrying out his job showed us our mistake. In the control room, as in the cockpit, users keep their eyes fixed on the screen for hours - the overall brightness of the work space is very dim and that of the tools also.

Design de strips

Two iterations of the design

Imprimante physique et virtuelle de strips

The two printers, side by side, the analog one and the digital one

As always, with the handling of the tool in mind, the design of certain elements (the strip printer or the strips themselves) related to work objects in their analogue form (skeuomorphism). Conversely, items related to digital data, such as datalink information, or comets representing the position of the aircraft in radar view were dealt with in a more "immaterial" mode, which could be defined as flat design.

Conclusion: experiments and results

A series of experiments was conducted in real-time simulation with controllers at the Paris control centre. The principle ideas of vertical display and the touch screen for input were validated and the input times found to be in the same range as those measured with the paper strip.

This result, obtained without prior training, was very promising. Moreover, tests on conflict retention following simulations confirmed that the vertical view did not deteriorate the mental representation of the situation. Some results even corresponded to a better representation (reduction by a half in level clearances, a doubling of speed clearances.)

During the presentation at the ATC Maastricht 2007 show visiting controllers testing the tool took ten minutes to handle it before starting to manage simulated traffic and then in a short time actually began to invent uses.

By adopting a "post-WIMP" (Windows, Icons, Menus, Pointing device) perspective, ASTER allowed the controller to perform expert actions using direct interaction. In the context of managing change the acceptance of this new tool is great advantage. Also, the fact of reaching a level of interface quality – from both a usage and aesthetic point of view – equal to that of a product has generated a sense of comfort, security and confidence in the machine, and has paid off. This has only been possible thanks to close collaboration within the research team, through the concept features, the development and the design.

Linked projects