EFIS - EICAS - man machine interface
Thread Starter
Joined: Jun 2003
Posts: 4
Likes: 0
From: cloud7
EFIS - EICAS - man machine interface
I AM INTERESTED IN ANY LINKS THAT COULD HELP ME TO DO AN EVALUATION OF A GLASS COCKPIT DESIGN FOR GENERAL AVIATION AIRCRAFT. THE EVALUATION AIMS AT HUMAN FACTORS / MAN MACHINE INTERFACE.
ARE THERE ANY KNOWN STANDARDIZED TEST/EVALUATION PROCEDURES?
ARE THERE ANY KNOWN STANDARDIZED TEST/EVALUATION PROCEDURES?
Joined: Oct 2000
Posts: 436
Likes: 0
From: N. Europe
First off, I'd suggest getting the caps lock off. Posting in ALL CAPS is hard to read and often considered the equivalent of yelling.
That aside, what aspects of the design do you want to evaluate and why? Compared to what? There has been much written on HMI/HFA. What is your background, what have you read so far and so on?
/Fred
That aside, what aspects of the design do you want to evaluate and why? Compared to what? There has been much written on HMI/HFA. What is your background, what have you read so far and so on?
/Fred
Joined: Dec 2001
Posts: 1,835
Likes: 3
From: Philadelphia PA
The FAA has published some good guidance material on this. Look on their website.
Otherwise, you have just entered into a real minefield where there is not a lot of hard guidance.
I would also suggest Dick Newman's books on the subject as a good starting place for how to evaluate these displays.
Otherwise, you have just entered into a real minefield where there is not a lot of hard guidance.
I would also suggest Dick Newman's books on the subject as a good starting place for how to evaluate these displays.
Joined: Dec 2002
Posts: 2,776
Likes: 350
From: UK
Start working with the design team immediately as any changes post design freeze may well be unaffordable. Avoid salespersons. Use several pilots for the assessment
If you do not think some thing is quite right first time; then it isn’t. Two or three exposures to a bad display can convince you that its OK – nothing has changed – it isn’t OK. Do not fool your self that you will learn to use a poor display.
In my experience, the design should be very simple / basic. Follow the principles of existing electromechanical instruments or a similar EFIS; think about issues of training or crews reverting to previously taught instrument interpretation when under stress. Form and understanding of how an existing or similar aircraft is operated then compare this with the intended EFIS operation. Always question what the purpose of each symbol is, what information is to be conveyed, when and where, and its relative importance.
Check for a consistent colour code, Airbus or Boeing; do not add colour just because it looks good. A consistent, logical operation is essential.
Refer to the principles of ‘Gestalt’ (a 20th century German psychological theory involving Proximity, Similarity, Closure, Connected, Continuity, and Closed Form); basically it means ‘If something looks alright then it is’. 'Alright' in this sense is aeronautically sound – for use in a cockpit and not on a PC screen. Thus your evaluation must be undertaken in the aircraft during airborne operations and should include normal and non-normal conditions. Check night lighting / reflections. Simulators give you a good feel for displays but remember there are just a big PC game and may not fully represent the real world.
The most important items in an EFIS are those which have been removed; i.e. not included due to weak argument as to their purpose not essential for the safety of flight. Do not add endless lists of information just because it can be done. Do not box items in because it looks tidy; ‘the minds eye’ tends to stick inside boxes (Gestalt theory) and increases the instrument scan workload. Be ruthless and de clutter everything – ‘blackspace’ contributes to the understanding of important information.
Just because a display works in another aircraft it may not work in yours. Beware of sensor and scaling issues; an Airbus lateral accelerometer (slip) display may not be, or is not the same as a GA slip ‘ball’ and any new EFIS certainly will require acceleration scaling specific to aircraft type. Do not use pure HUD symbology.
Obtain an understanding of the use of sensors and source information; how the data is ‘smoothed’ or mixed with other data e.g. an airspeed trend vector may use a form of inertial smoothing, if this is flight path dependent then oddities may occur during stalls or upset manoeuvres – speed decreasing but with speed trend (flight path) accelerating – misleading?
Maps are very powerful displays; again do not clutter the display and be very careful to use only reliable (certificated) navigational sources. Use raw data VOR / DME for overlays on FMS data – then you can visually check for map shifts. Use large clear switch selections; avoid push buttons and menus. Menus up to four deep are increasingly poor, more than four … NEVER! Annunciate important selections / modes both at the point of selection (switch) and at the point of operation (display).
Many other pitfalls, too many to recall here. EFIS assessment is demanding, rewarding, and can be great fun. Good luck.
If you do not think some thing is quite right first time; then it isn’t. Two or three exposures to a bad display can convince you that its OK – nothing has changed – it isn’t OK. Do not fool your self that you will learn to use a poor display.
In my experience, the design should be very simple / basic. Follow the principles of existing electromechanical instruments or a similar EFIS; think about issues of training or crews reverting to previously taught instrument interpretation when under stress. Form and understanding of how an existing or similar aircraft is operated then compare this with the intended EFIS operation. Always question what the purpose of each symbol is, what information is to be conveyed, when and where, and its relative importance.
Check for a consistent colour code, Airbus or Boeing; do not add colour just because it looks good. A consistent, logical operation is essential.
Refer to the principles of ‘Gestalt’ (a 20th century German psychological theory involving Proximity, Similarity, Closure, Connected, Continuity, and Closed Form); basically it means ‘If something looks alright then it is’. 'Alright' in this sense is aeronautically sound – for use in a cockpit and not on a PC screen. Thus your evaluation must be undertaken in the aircraft during airborne operations and should include normal and non-normal conditions. Check night lighting / reflections. Simulators give you a good feel for displays but remember there are just a big PC game and may not fully represent the real world.
The most important items in an EFIS are those which have been removed; i.e. not included due to weak argument as to their purpose not essential for the safety of flight. Do not add endless lists of information just because it can be done. Do not box items in because it looks tidy; ‘the minds eye’ tends to stick inside boxes (Gestalt theory) and increases the instrument scan workload. Be ruthless and de clutter everything – ‘blackspace’ contributes to the understanding of important information.
Just because a display works in another aircraft it may not work in yours. Beware of sensor and scaling issues; an Airbus lateral accelerometer (slip) display may not be, or is not the same as a GA slip ‘ball’ and any new EFIS certainly will require acceleration scaling specific to aircraft type. Do not use pure HUD symbology.
Obtain an understanding of the use of sensors and source information; how the data is ‘smoothed’ or mixed with other data e.g. an airspeed trend vector may use a form of inertial smoothing, if this is flight path dependent then oddities may occur during stalls or upset manoeuvres – speed decreasing but with speed trend (flight path) accelerating – misleading?
Maps are very powerful displays; again do not clutter the display and be very careful to use only reliable (certificated) navigational sources. Use raw data VOR / DME for overlays on FMS data – then you can visually check for map shifts. Use large clear switch selections; avoid push buttons and menus. Menus up to four deep are increasingly poor, more than four … NEVER! Annunciate important selections / modes both at the point of selection (switch) and at the point of operation (display).
Many other pitfalls, too many to recall here. EFIS assessment is demanding, rewarding, and can be great fun. Good luck.
