The sociotechnological view of industrial safety
Résumé
As part of the European project STARS (Socio-Technological Assessment in Risk Regulation regimeS), this presentation wishes to come back on the meaning of a ‘sociotechnological’ view of industrial safety. In March 2011, a tsunami in Japan hit and passed over a wall that was designed to protect the nuclear power plant located behind it. But the wall was too low. Water flooded the premises where backup batteries were located, and a loss of electric power occurred as a consequence. Without power, the process operators, engineers, managers, regulators, media, officials and civil society act in face of the behaviour of a complex system that they have never experienced under these specific circumstances. Meltdown of the core of the reactor ensued. Thousands of people were evacuated and the environment contaminated. In April 2010, highly flammable pressurised gases from deep geological layers shot up to an offshore platform, forming a cloud which ignited and exploded. The measures designed to prevent such an event, including a blow out preventer (BOP), did not function as intended. The purpose of such a device was to cut the line connecting the offshore platform to the well in order to stop the gas from reaching the platform. Several people were killed, the platform sunk, and a pollution followed. Operators, engineers, managers, authorities, regulators, shareholders, media and civil society faced the biggest oil pollution of US history. In June 2009, sensors indicating speed in the Air France aircraft flying from Rio to Paris got frozen and sent wrong indications to pilots. The aircraft was going through a well known storm area. Manual handling of flight by pilots in these circumstances is then impaired by interface design issues conveying contradictory information to pilots about aircraft’s status. The crash left no survivors and became the deadliest crash in the company’s history. To make sense of such events, disaster and safety research has been steadily producing a diversity of concepts, models and studies over the past thirty to forty years from a diversity of disciplinary backgrounds (psychology, ergonomics, management, sociology) and safety-critical systems (aviation, nuclear, railways, chemical). As a consequence, one encounters a range of ‘research traditions’ and key topics mixing together graphical, metaphorical and analytical insights. Safety Culture, Safety Management, High Reliability Organisations, Resilience Engineering, Swiss Cheese Model, are some of this most visible examples which have been framing through specific lenses the issue of expanding our understanding of safety beyond a technologically centred approach. The contention of this paper is that, beyond and throughout this diversity, one can nevertheless extract a consistent background of works which attempt to conceptualise the sociotechnological side to safety by incorporating under this heading a vision interweaving technology, tasks, activities, processes, structure, cognition, power and culture. Based on debates anchored in social theory (among which the micro-macro issue or the materialistic side to society) and interdisciplinary studies (among which the drive for convergence), this move helps to shape a new generation of safety models that take stock of the managerial, social and political dimensions in our effort to understand and prevent disasters in a world of accelerating and globalised changes. The presentation will introduce the main arguments supporting this sociotechnological thread.