The Fiction of Function Allocation

FUNCTION ALLOCATION IS THE process that assigns work roles and their tasks to the humans and machines in a system. Or so the textbooks are written for students of human factors in design. Function allocation, since its emergence as a concept 40 years ago (see inset on page 23), has become part of the conventional wisdom of the human factors profession and, in turn, that of many federally regulated and military-industrial design communities. Function allocation has led to more systematic input of human factors to the design process and to the improved design of large-scale, complex, dynamic sytems. Well maybe. Certainly function alloca, By Robert B. F u I d tion has been an undeniable success as an explanatory framework. A body of literature, textbooks, and regulatory guidelines can be cited that explain function allocation and its performance as a component of design. Numerous articles on process control, automation, and flight deck performance suggest that an inadequate allocation process contributes to degraded system performance. (Blaming the operator may be antithetical, but blaming the designer is de rigueur.) The allocation issue has helped to create a mandate for human factors participation in the early, conceptual stages of complex systems development. But there's scant evidence anywhere that the practice of formal function allocation has done much to improve the design of large-scale, complex systems. Once I, too, was a whole-hearted advocate of the allocation process. Over time, however, I've come to suspect that allocation is an artifact of academic explanation that has yet to be validated as a practical process. This position can be supported by