This article interrogates an aircraft control system referred to as 'fly-by-wire'. First developed in the 1960s, fly-by-wire replaced mechanical and hydraulic aircraft control systems with an electronic, computer-mediated system capable of relaying, responding to, and sometimes restricting, human inputs from pilots. In so doing, fly-by-wire enabled an entirely new world of flight in which human decisions were subject to machinic, and electronic, approval. The article examines the effects of

more »

... by-wire on the socio-technical control of aircraft, with repercussions for how one considers contemporary questions regarding the interweaving of automation, control, knowledge and safety. It proceeds in two parts. Firstly, it argues that fly-by-wire is a form of 'distributed control'. Dependent upon situated automation, the ability to control, steer and manoeuvre the aircraft is variously distributed beyond the cockpit and human pilots, to integrated components, sensors, physical surfaces, and systems throughout the aircraft itself. In so doing, new and novel operational capacities are reached depending on the situation; shifting and re-calibrating the relationship between pilots and aircraft. Secondly, and more specifically, I suggest that distributed forms of control in the shape of aircraft fly-by-wire systems yield so-called 'automation surprises'. The effect of distributing decision-making to a wider assemblage of components, sensors, surfaces, and systems is that operational asymmetries occur in the otherwise smooth collaboration between pilot and machine. I discuss recent Boeing 737 accidents in order to evidence this argument, contending that recent additions to fly- by-wire have led to novel re-distributive control effects. As the development of prototype autonomous vehicles abounds, historical lessons drawn from aircraft control, decision-making and safety should be of critical importance.