Designing with consideration of the human factor: Changing the paradigm for higher safety

Aerospace avionics and operation need another dimension. The “glass cockpit” came as a response to an impossible perceptive situation, when pilots had to monitor countless gauges (the concept being “one info-one gauge”). The change came with the renegotiation of the info/surface ratio given to data at the pilot level by using CRTs to integrate data in a user-friendly (dedicated) format. Since then improvements were made on the same paradigm, through very efficient cyclic design. Yet, in a way and to be blunt, little has been done, but just cosmetic changes regarding the distribution and size of screens. One does not want to change a ”winning team”. But the efficiency of the paradigm has faded away with the evolution of the aeronautical environment (traffic increase & permanence of service). The today's problem lies with “non-defective aircraft” monitored by “perfectly trained crews” still falling from the sky. One explanation is, at the crew level, that we have reached a system complexity that, while acceptable in normal conditions, is hardly compatible with human cognitive abilities in degraded conditions. The today's mitigation of such risk relies on the enforcement through intensive training to manage extremely rare (off-normal) situations explained by the potential combination of failures of highly complex systems with variable environment & with variable humans. Looking back into the limits and strengths of operators, we may find with some very basic knowledge on human cognitive strategies ways to revisit and review our design principles to give back to pilots the ability to stay in the loop: not through the management of more & more complex systems, but by helping them doing what they do best, manage their own resources to make adequate decisions. First, basic ground rules regarding human factors must be known and accepted by the designers and planners. Second, design must respect cognitive strategies “naturally” followed by operators to spare their resources. A form of “ecological” design will follow, facilitating operators' economical interactions. In my opinion, the probabilistic design for reliability concept is an important step in the right direction.