Man and Machine. The Modern Cockpit
Human-machine interaction is an important consideration in the development of cockpit technologies. A.R. Prince examines the issue.
The modern helicopter cockpit appears much simpler than that of earlier generations with fewer discrete instruments. This is because the cockpit has been automated and automation in aviation has been introduced specifically for the purpose of simplification. The highest level of automation attained in the civil sphere is in the large jetliner, but unfortunately, while automation has improved safety, it has also been a factor in accidents.
The underlying difficulty is the interaction between the physical and the non-physical. The physical is the cockpit and its systems, the non-physical the pilot's cognition. The pilot is modeled in cockpit design through how it is considered he/she will perceive sensed information, comprehend it, and then act. The required sensed information in earlier years of aviation was a limited amount and so it was easy for the pilot to be aware of all important aircraft events. The pilot had what is called 'situational awareness' and this enabled the pilot to operate the aircraft with a proficiency to ensure survival.
The pilot’s sense information was obtained from the scanning the various electromechanical displays and from looking out of the window. But as the aircraft evolved, sensors and their displays began to proliferate and the increasing volume of presented information became challenging. However instrument technology was also evolving. It began to allow the condensing of information. The technology miniaturized electronics and the displays became the electronic. The information processing capability of the electronic display could and soon did go beyond just the condensing of information. It became multifunctional with the information processing happening without the pilot's knowing. The result was the pilot’s situational awareness began to decrease.
The modern systems manufacturer still seeks to improve the capability of electronic displays but it is a challenge is to do so while ensuring the pilot retains situational awareness. The most capable display a manufacturer has yet developed is one which covers the whole of the instrument panel. It is used in the Bell V-280 military tiltrotor, which is due to fly in 2017. This display will be discussed in more detail below. Another concept based on a single display is from the French company Thales. It has led to the four-display Avionics 2020 concept, for a variety of aircraft types including helicopters of any size. The helicopter versions would be scaled-down jetliner versions.
Thales' ability to scale displays suggests that it has standardized information presentation. This is encouraging. The more standardized the displays, the more definitive can be the pilot's own methodology for interacting with them and with the cockpit.
The industry has also realized that an advertised characteristic of automated systems bears re-examination. This is the reduction of pilot workload. A high workload wasactually what the early pilot faced. The checking and re-checking of multiple displays was certainly tedious but it was what enabled the pilot to gain and retain situational awareness.
One of the most notable automation-related accidentswas that of an Air France Airbus A330 jetliner in the Atlantic Ocean in 2009. All onboard died. Just prior to the crash the pilots were not aware of what the system was doing. It had malfunctioned and handed over control to them but they did not know the underlying system logic. In their lack of awareness they incorrectly controlled the aircraft. Their experience is in sympathy with that of many pilots in similar situations, confusion about what is happening and the expected future system actions and system logic.
From these experiences and greater understanding of human-machine interaction, industry is aware that confusion will occur when a complex logic-based system is left to operate on its own without first there being operator understanding of the system and its operation feedback to the pilot. Pilot workload cannot be eliminated, for even if there is an automated system at work, it will need oversight. And even for the pilot to oversee it, the pilot must first know it and how it functions. Its function would not be a mystery and as result, the pilot would be the true master of the system, and be able to fully employ its capabilities.
System capabilities have increased as the automated system has become more advanced, but increased capability in turn means that there is more about it
that the pilot must learn to master it. Notably pilot errors with such systems, as occurred in the A330 incident, have tended to be more cognitive,
with less of the motor skill aspect as seen with non- or less automated systems. For example, the A330 system has multiple modes of operation and the
pilot confusion seen in the crash was not just about what the automatic system was doing, but which mode it was
operating in; and when the mode would change and how the new mode would operate. Industry has realized that the problem with multi-modes and otherwise
complex systems is that the more complex the system, the more likely there would be unknown errors in the system. Again, this occurred in the A330
incident. With systems of such complexity the greater is the requirement for knowledgeable-pilot oversight.
Multiple modes are found in flight control systems of the fly-by-wire (FBW) aircraft. The A330's is an example. The A330 and other experiences involving
loss of pilot situational awareness have been instructive for aircraft designers. They are realizing that as system autonomy increases concomitantly
they have to ensure pilot situational awareness. The group of aircraft designers includes those of helicopters. They have now developed the West's
first FBW civil helicopter. This is the 16-passenger Bell 525, which first flew in July 2015. The more recent V-280 also has a FBW system, and as the
V-280 is more complex than the 525 due to being a tiltrotor, its FBW system is probably more advanced. HeliOps spoke to a Bell lead engineer on the
V-280 program. He affirmed that it is the company view that the pilot should be aware of what the system is doing. He said that in the case of the
V-280, the system, "makes it absolutely clear to the flight crew," what it is doing. This would be through indications visually, on the display, and
In FBW systems the traditional mechanical-based control system is replaced with an electronic-based one. In this system pilot control inputs at the collective and cyclic controls and tail-rotor pedals are no longer just transmitted to the actuators; system computers first interpret them. FBW systems are indeed more complex than mechanical ones. The 525's though will most certainly incorporate a lower level of multi-modality than the A330's. Still, FBW-system complexity compared to mechanical-based flight control systems is primarily why a comparatively large helicopter was chosen, and why it is the only civil one having or envisaged to have FBW flight controls.
Another area the designers of highly automated helicopters have to consider is the loss of vehicle controllability due to what is called rotorcraft-pilot couplings. The latter is an unbalanced force couple about a helicopter axis, caused in the simplest case by pilot control action. Balance may be restored by removal of the control action. The problem is when the helicopter is highly automated because it is then particularly susceptible to another type of coupling, due again to pilot control action. The control action in this case, because of the automated system's feedback loop, may lead to dynamic unbalancing, that is, helicopter oscillation about the axis. Danger arises because pilot removal of the control action may not be possible. Dynamic unbalancing would be more of a concern for the tiltrotor than the conventional helicopter, since it has off-axis thrust provision, and in-flight thrust-vectoring. So Bell would have considered rotorcraft-pilot couplings in the development of the V-280. It would have benefitted from its experience with the in-service V-22 military tiltrotor.
Another manufacturer exploring autonomous helicopter capability is Sikorsky. In January 2016 a company pilot took off, flew and landed a S-76D helicopter without the use of flight controls. Instead he used a touchscreen hand-held computer. Touchscreen technology has been employed on fighter aircraft for several years. More recently it has become ubiquitous in consumer electronics. It is now also entering the civil helicopter sphere. The two-pilot Bell 525 has four Garmin touchscreens and Garmin technology is also available for other helicopters. The concept is also the basis of the Thales Avionics 2020 display, and in fact what makes it scalable to different cockpits.
The touchscreen is the latest product in the trend to condense cockpit information. Control action and system response occur in the same area. This allows
feedback to be immediate; flight control is more responsive. Further, the screen area will naturally be larger than for previous displays since the
buttons and keys previously on the side of or underneath can be reduced in number or removed altogether. Designers could also be tempted to further
enlarge the display for more information and improve tactile-control. With regard to the former in particular, the V-280 mockup has demonstrated underlaying
backdrops of simulated daylight, day synthetic-vision, and low-light-amplification night views of the world, obtained from aircraft sensors. The V-280
represents the level of technology Bell envisages could be in service in 10-15 years. HeliOps asked the Bell engineer what, in V-280 simulator studies
so far, was the most popular way pilots chose to configure the display. In his reply he indicated that how a pilot, “integrates information,” will
determine how he wants to receive it on the display. This statement indicates increased company and industry recognition of the importance of pilot
control and situational awareness. He further indicated that a lot of pilots in day visual flight rules conditions, “turned the system down,” not wanting
display information to distract them from the situation outside the cockpit. He said Bell is learning as it develops the system. The feeling at the
company perhaps could be likened to that of being on the frontier of the understanding of pilot-cockpit linking.
If the pilot considers that the backdrop can contribute to confusion, he could remove it altogether. Generally a pilot using touchscreens would be able to reduce display density as he chooses, and thus could reduce his workload. Of course he would not want to reduce it to the extent that situational awareness erodes. If the touch technology is based on a technique now established in consumer electronics, the pilot might also note usability differences compared to the consumer electronics. Specifically, a firmer touch may be required, to prevent unintentional inputs. In the case of the 525 though, this is not important. The screen has an infra-red grid across it, which the pilot's finger disrupts.
The V-280 also gathers more external information through a new type of system and this information could be shown on the display. The system comprises sensors distributed on the vehicle-body. HeliOps asked the Bell engineer if such sensors could migrate to commercial helicopters. He said they would be useful to them. He specifically mentioned emergency medical services, among the most dangerous of civil helicopter roles.
In its S-76D experiment Sikorsky used the touchscreen in place of traditional controls. The latter are the collective and cyclic and the tailrotor pedals. Bell has replaced the first two in the Bell 525 with sidestick controllers. These could be more reliable, and would probably also be more comfortable to use and allow more space between the pilots and the instrument panel. FBW sidestick controllers though will not provide the pilot with tactile response, which mechanical-based systems do. Such a response is important to inform him of aircraft flight status and of control power still available. Some fixed-wing FBW aircraft have aircraft control response fed back to the sidesticks, and this could be a model for future FBW rotorcraft. Still a feature that Bell does have on the 525 sidesticks, is a mechanical connection between the pair, allowing either of the pilot and co-pilot to be informed if a control action is taken by the other.
The trends in helicopter cockpits are seen to include increased automation, most particularly, increased information concentration, on the path towards the unitary touchscreen display. New systems are bringing increased capabilities that in turn require more interaction from the pilot.
Automated systems though are evolving and will continue to evolve to increasingly consider pilot usage. Such a system will monitor multitudinous parameters, noting any unwanted changes as they arise, and alerting the pilot by aural or visual signals. This will become more central in its development while increasingly being in the context of what the human is good at. The pilot would then have the ability, being well informed and well trained, to act on provided information. It is not so much that the automated system will take autonomous action. R ather the pilot, knowing how the automated system operates, and being practiced in acquiring knowledge of the aircraft in the environment and controlling it whatever the system state, will be the one in command. That is when the automated system, the heart of cockpit technology, can be said to have matured. The pilot will undoubtedly then be the master of the machine.