Completed Projects

Funded by DFG

Increasing air traffic raises the requirements on future flight trajectories coupled with the necessity to follow a flight path with higher precision. Those requirements also have to be fulfilled in manual flight. This leads to an increased pilot workload especially in order to command all control devices that are changing the state of the total energy (potential, kinetic energy), i.e. thrust, speedbrakes and others. Thus, the proposed project aims to assist the pilots with an innovative longitudinal acceleration control command system (n_x control) that combines the command to increase or reduce energy in a demand value for a single inceptor. The functionality of the n_x control command system has to be comprehensible and transparent to the pilots in order to ensure an adequate situational awareness for monitoring and operating the controller, to increase the pilots’ acceptance, and to minimize training effort. To achieve this, the new controller has to be developed based on the prevalent mental model of the pilots regarding the control of the aircraft in longitudinal direction. Thus, the development of the n_x control function shall be based on a pilot-centered design approach and shall comprise the realization of the controller as well as the human-machine interface (display and inceptor). Although recommended, such new concepts are rarely investigated and validated from both the technical and the psychological point of view.

Consistent with the pilot-centered design approach, the prevalent mental model of the pilots regarding flight path control of the aircraft in longitudinal direction will be determined. That information is used to define the characteristics and the functionality of the n_x controller in detail. Furthermore, the effects of the higher degree of automation on pilots induced by the new controller and the resulting consequences for the concept and design of the human-machine interface will be investigated. Then, the n_x control augmentation system that calculates the command values for the energy-related control devices is designed and implemented. Simultaneously, a concept for the human-machine interface for the cockpit adapted to n_x control will be designed. Both controller and human-machine interface will be integrated into a flight simulator for proof of concept. Simulator test campaigns with experienced airline pilots serve to confirm the hypotheses that n_x control (1) leads to a higher precision in manual flight regarding preset and complex flight paths, (2) does not add pilot workload due to higher complexity and (3) has no negative impact on the situational awareness due to the higher degree of automation.

Contact persons: Prof. Dr. Dietrich Manzey [1], Dipl.-Ing. Simon Müller

Funded by DFG as part of the Research Training Group prometei [2]

The design of alarms systems and warning strategies in complex human-machine-systems requires a good understanding of the behavioural impact of alarms on human behaviour. Theoretically the handling of alarms can be described as a complex decision-making problem that can be formalised using signal detection theory. Such approach primarily focuses on the impact of validity of alarms on operator behaviour, i.e. the operators' trust in the systems and how operators response to it. Recent research suggests that two different kinds of responses to alarms might be distinguished which have been referred to as "reliance" and "compliance" (Meyer, 2004). Reliance reflects how much operators trust in the alarm function, i.e. that they get an alarm in case of critical system states. This behavioural tendency should mainly be determined by the rate of "misses" of the system. Compliance reflects the extent to which operators really follow the alarm, i.e. trust that there really is a critical system state in case of alarms. This tendency should mainly be determined by the rate of "false alarms". Research in this area includes experimental studies focussing on the impact of different aspects of alarm sensitivity and validity on the reliance and compliance of operators in handling the alarms. In addition, also other possible factors affecting the operator responses to alarms are considered (e.g. base rates of critical events; operator workload). 


Contact: Magali Balaud [3]

Within the framework of the BMWi reactor safety research program, we are studying a previously unrecognized element of safety culture, i.e. the significance of shared knowledge structures in high-hazard organizations for ensuring safe operations. Although the influence of shared knowledge on behavior is indisputable, this factor and especially its impact on safety culture have been widely disregarded in research on the determinants of safety in high-reliability organizations. Relevant background theory is provided by concepts developed in social-cognition research, which has often been limited in focus on the possibilities of improving performance in small teams through better team processes. Within the current project, we will analyze how shared knowledge structures not only emerge within small teams, but also within organizations, and what kind of positive and negative effects on behaviour can become apparent. For this purpose, field studies as well as laboratory experiments are being conducted. A large field study in a German nuclear power plant will focus on the existing pattern of information distribution in a high-reliability organization and its impact on establishing shared knowledge. In addition, laboratory experiments are being performed in order to study specific aspects of shared knowledge as a determinant of individual behaviour in teams. This includes studies on the effects of human redundancy on system operations where shared knowledge is usually regarded an important means to enhance safety.
Contact: Juliane Marold