Nowadays automation is used in the cockpit primarily for technical reasons – what the pilots are doing is rarely considered. As accident investigations have shown, problems with the human-machine interface contribute to 60 to 80 per cent of all accidents. The newly developed A-PiMod cockpit architecture (Applying Pilot Models for Safer Aircraft) improves the cooperation between human and machine by not only monitoring the condition of the aircraft and the ambient conditions, but also the state of the pilots. On the basis of the pilots’ eye movements, gestures, and entries, conclusions are drawn about their current intentions, their awareness of the situation, and their workload in order to support the pilots as well as possible in any given situation. In addition, the cockpit architecture that was developed enables a dynamic distribution of tasks between crew and machine. The tasks at hand are identified in the current situation and subsequently all task assignments and the related risks of error are determined. To faciliQ2:Q6tate cooperation between humans and systems, in A-PiMod conventional interaction in the cockpit is enhanced by speech recognition software and touchscreens.
With this approach, the European partners of the A-PiMod project are helping make flight traffic safer thanks to improved human-machine interaction in the cockpit.