Design of Natural Warning Sounds in Human-Machine Systems

Detta är en avhandling från Stockholm : KTH

Sammanfattning: The goal of this research is increased safety in aviation. Aviation is a highly automated and complex, as well as, safety critical human-machine system. The pilot communicates with the system via a human-machine interface in cockpit. In an alerting situation this interface is in part an auditory alerting system. Human errors are often consequences of actions brought about by poor design. Pilots complain that they may be both disturbed and annoyed of alerts, which may affect performance, especially in non-normal situations when the mental workload is high. This research is based on theories in ergonomics and cognitive engineering with the assumption that improved human performance within a system increase safety. Cognitive engineering is a design philosophy for reducing the effort required by cognitive functions by changing the technical interface, which may lead to improved performance. Knowledge of human abilities and limitations and multidisciplinary interrelated theories between humans, sounds and warnings are used. Several methods are involved in this research, such as literature studies, field studies, controlled experiments and simulations with pilots. This research defines design requirements for sounds appropriate in auditory alerts as Natural Warning Sounds. For example, they have a natural meaning within the user’s context, are compatible with the auditory information process, are pleasant to listen to (not annoying), are easy to learn and are clearly audible. A design process for auditory alerting systems is suggested. It includes methods of associability and sound imagery, which develop Natural Warning Sounds, and combines these with an appropriate presentation format. Associability is introduced and represents the required effort to associate sounds to their assigned alert function meaning. An associable sound requires less effort and fewer cognitive resources. Soundimagary is used to develop sound images. A sound image is a sound, which by its acoustics characteristics has a particular meaning to someone without prior training in a certain context. Simulations of presentation formats resulted in recommendations for cancellation capabilities and avoiding continuously repeated alerts. This research brings related theories closer to practice and demonstrates general methods that will allow designers, together with the users of the system, to apply them in their own system.

  KLICKA HÄR FÖR ATT SE AVHANDLINGEN I FULLTEXT. (PDF-format)