PPRuNe Forums - View Single Post - Man-machine interface and anomalies
View Single Post
Old 16th Apr 2012, 02:34
  #43 (permalink)  
Lyman
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes on 0 Posts
Grrr

Hi safetypee.

Fluid, Additive, Synergistic, and Timely. Call it FAST. Aviation brooks no timeouts. Thinking is time consuming, one reason why auto is superior in rote and program over human abstraction. Assessment is a pedestrian way to address not 'risk' but risk management. Auto or no auto, without a highly developed intuiton, pilots can create, rather than minimize, risk. Here I use the word to stand in for the ideal pilot, experienced, highly trained, and confident.

447 had interface problems. I don't fault the system, nor do I fault the pilots, per se. There may never be a satisfactory conclusion, but the advances have begun.

"First, do no harm." None of the more obvious blunders are outside the realm of human error. Autopilot is not a computer, it is a system of servos, acting on behalf of. Understanding the Flight control logic is incumbent upon the crew, not vice versa. What new issues? I called the interface a "partnership" for a reason. Since the beginning, pilots of renown are the ones at one with the machine.

Lack of AoA indication, Difficulty in seeing the partner's SS, Faulty probes, not being fluent in other than Normal Law; these are not interface issues, but instead, bonehead mistakes in design, programming, cockpit layout, and poor training. When plugging the dike, eleven holes exceeds the available plugs by one, and one open hole is enough, let alone several.

The successful pilot will master the machine, but has degraded chance for success with inexcusable lapses in design, training, consistency, etc.
Lyman is offline