Human-machine interface extends far beyond the flight deck. As well as pilots in cockpits there are controllers separating and guiding aircraft.
The recognition and development of HMI in this area parallels the evolution of the aircraft HMI. In each case the evolution is from analogue to digital systems, with the trend being to more information and more integration of information. Where Aeradio network operators, established in Australia by AWA for the Department of Civil Aviation in the 1930s, predecessors of today’s air traffic controllers, had to carry a picture of the (admittedly sparse) air traffic in their heads, the modern air traffic controller sees it displayed on a screen, sometimes with potential conflicts highlighted.
Another discernable trend is towards design for ease of use. In the era when aircraft flight decks consisted of a few brass-ringed instruments mounted on a wooden panel, Aeradio stations consisted of powerful but primitive valve transmitters, Morse keys and direction finding equipment, often arranged ad hoc on an everyday table or desk.
Limitations of technology meant humans were the ones that adapted. The Morse code key was an example of this. Communication could only be carried out between people who had learned the sequence of dots and dashes.
The Aeradio operator of the 1930s had six sources of information: Morse transmissions from aircraft and other stations, air-to-ground voice radio at shorter ranges, the telephone, the local meteorology officer—who was often located in the room next door, and, finally, his ears. It was relatively common, in overcast conditions, for the radio operator to step outside and listen for the sound of aircraft engines. The operator would then radio the aircraft to say it had passed over the aerodrome, or could be heard in a particular direction.
‘I think the people who learned on the strip system developed a more fundamental understanding of the airspace. You had to think for yourself in four dimensions.’
As instrument flight became routine, then mandated, in post-war aviation, the controller’s station evolved in parallel. The flight strip system was developed, reportedly, from a system used in the US for managing railway wagons in marshalling yards. Details of each flight in a sector were written on a paper strip that was moved up and down a column of strips depending on when the next action was due, and deleted when it had cleared the operator’s sector. The strip system established a foundation of safety: orderly, procedural workflow.
Dedicated consoles appeared in the early 1960s and incorporated expandable modular design, battery back-up, duplicate audio systems and ergonomic design, in the form of curved map displays, and lighting.
On some ATC consoles flight strips could be put into slots which had interlocks to indicate assigned altitude on other consoles by means of coloured lights. This guarded against multiple aircraft being assigned the same altitude.
‘It sounds arcane and complex, but it worked,’ says vice president of the Civil Aviation Historical Society, Phil Vabre. ‘I think the people who learned on the strip system developed a more fundamental understanding of the airspace. You had to think for yourself in four dimensions.’
Over the history of ATC the mental airspace picture in the operator’s head has become visible, through technologies including several generations of radar, and, most recently, ADS-B.
‘Early primary radar did not identify one aircraft from another,’ Vabre says. Instead controllers would ask aircraft to perform tell-tale manoeuvres for identification, (“Alpha Bravo Charlie, for identification turn left heading …”) after which the controller would stick plastic identifiers known as ‘shrimp boats’ to the screen above a return. Bumping the console reset the identification system, Vabre drily notes.
Electronic screen marking first appeared in the form of an inter-console marker which highlighted a return for hand-off to other controllers.
Society president, Roger Meyer, remembers, ‘You sometimes had to wiggle it around to get attention. But it was a positive handover, which was good.’’
By the late 1980s, digital radar processors were able to convert raw radar returns into display data, overlaid on a map. Call sign, altitude, speed and track could now be displayed before the controller.
‘The radar data processor basically decided what was an aeroplane and what wasn’t,’ says Vabre.
1994 saw the introduction of The Advanced Australian Air Traffic System (TAAATS), which added a flight data processor into the system. TAAATS fused multiple data sources into a single picture,’ says Vabre.
‘It created a four-dimensional trajectory for the aircraft and put that on a situational display, including the tracks of aircraft that weren’t under surveillance coverage. Previously a procedural board had been used to do that. One lot of tracks was generated from radar, the other lot from flight plans but they were all shown on the same screen.
‘On top was radar data, if that wasn’t available there was ADS-C, (automatic dependent surveillance—contract a satellite datalink system) and finally flight plan data. There were display rules to show what the data source was, because different data sources had different accuracy—the flight plan was only where the aircraft intended to be, for example.’
TAAATS is due to be replaced, along with the defence air traffic system ADATS, by OneSky, an integrated military and civilian ATC system. OneSky is to be delivered by Thales Australia, local subsidiary of French aerospace, defence and transport conglomerate Thales.
Thales Australia has a strong focus on HMI research and design. Thales ATC design specialist Grant Williams says, ‘Melbourne has become the centre of excellence in the organisation for HMI. The new HMI was designed and developed principally here.
‘We’re investigating new technologies coming in on the IT side, eye tracking, mobile devices, gesture control and voice recognition. This is looking into the future but we need to start looking now and some we’ve already incorporated into our products.’
OneSky, which will begin rolling out in 2018, will be based on the Melbourne-developed Thales TopSky-ATC system, already in service in 130 locations around the world.
Thales says TopSky embraces ‘a minimalist display philosophy for easy and quick comprehension of the air situation.’ Flight information is sparse in the standard display, but further details are instantly available with a mouse click roll of the scroll wheel. Track information comes from many surveillance sources, including Mode-S, secondary radar and ADS-B. A powerful processor enables advanced track conflict detection, which is also incorporated into the track display.
Thales ATC specialist and former controller Geoff Bates says simplicity and ease of use are the guiding principles of TopSky. ‘In the new HMI we’re focused on decluttering and only giving information that’s important, but making it efficient to get more information if required.’
‘For example, tracks cruising at their assigned level only have a couple of pieces of information on display. As soon as they’re climbing or descending, the label becomes a bit bigger, with more items—a simple visual cue to the operator an aircraft is maneuvering.
‘Users can set up their own workspace, and decide the number of parameters they want on their aircraft. They can save multiple versions of their preferences so if they work different airspace sectors on different shifts they can log on and choose the appropriate version.’
There’s a lot of use made of colour, so that users can very quickly get the correct status of the aircraft. Green is a standard colour that means everything’s OK. Yellow indicates a caution, such as flight plan conflict and if I see red I really need to pay attention.’
‘The colours are suitable for a well-lit daylight room, because we’ve moved out of the dark caves of the old centres.’
Williams says there is also a generational aspect to acceptance of new designs.
‘We had similar issues in the early days of TAAATS—controllers who were not familiar with a mouse. Now we’re finding with the new HMI the younger controllers are jumping in and embracing the concept.’
‘At the end of the day you are still relying on what’s in the controller’s head. All these systems have been designed to support that, rather than try to take over.’
CASA executive manager aerodrome and airspace regulation, Peter Cromarty, says ATC automation gets it right when it acts as a support, rather than a substitute for decision making.
‘During the 90s the Eurocontrol Experimental Centre at Bretigny near Paris conducted many simulations into tools that helped controllers. They were pictorial, very intuitive and allowed much quicker decision making by the controller. The idea was to allow the controller to do what he liked to do—wheel/deal and take decisions, and allow the system to do what it was good at—monitor the situation. It’s the obverse of the path taken on flight decks.’
Cromarty was impressed by how easily controllers were able to learn the experimental system. ‘The FAA sent over five controllers to take part. Within a few hours, and without previously knowing the airspace, they were able to control twice the traffic that the then controllers could handle on the actual sectors on a busy day.’
Vabre, who is also a controller, says despite technology, the basic job and skills of the air traffic controller are unchanged. ‘At the end of the day you are still relying on what’s in the controller’s head. All these systems have been designed to support that, rather than try to take over.’