Function allocation for humans and automation in the context of team dynamics

Abstract Within Human Factors Engineering, a decision-making process called function allocation (FA) is used during the design life cycle of complex systems to distribute the system functions, typically identified through a functional requirements analysis, to all human and automated machine agents (or teammates) involved in controlling the system. Most FA methods make allocation decisions primarily by comparing the capabilities of humans and automation, and then by considering secondary factors such as cost, regulations, and the health and safety of workers. The primary analysis of the strengths and weaknesses of humans and machines, however, is almost always considered in terms of individual human or machine capabilities. Yet, FA is fundamentally about teamwork in that the goal of the FA decision-making process is to determine the optimal allocations of functions among agents. Given this framing of FA, and the increasing use of and sophistication of automation, there are two related social psychological issues that current FA methods need to address more thoroughly. First, many principles for effective human teamwork are not considered as central decision points or in the iterative hypothesis and testing phase in most FA methods, despite the fact that social factors have numerous positive and negative effects on individual and team capabilities. Second, social psychological factors affecting team performance can be difficult to translate to automated agents, and most FA methods currently do not account for this effect. The implications for these issues are discussed.