New tools, old tasks: Understanding the implications of new technologies for the safety of sociotechnical systems – the case of Integrated Operations

Abstract

Integrated Operations (IO) denotes an operating regime in the petroleum industry where new technologies and work processes make possible an increased use of real-time data, collaboration across disciplines and geographical distances, and expert knowledge, with the goal of achieving better, faster and safer operations. This thesis is based on a study of drilling operations and the way in which IO may affect the safety of the operations. The objective of the study is to explore the theoretical and methodical conditions for understanding sociotechnical systems and change processes, and to investigate how the conditions for safety in sociotechnical work is influenced by new IO technologies. A sociotechnical perspective inspired by science and technology studies and Actor- Network Theory has guided this work both in the fieldwork phase and in the theory construction. Observations and interviews with professionals from oil companies, service companies and technology providers constitute the major sources of the empirical material.

Specific research questions have been derived from the main objective, and these research questions are explicitly addressed in the four journal articles included in this thesis. While the empirical investigations and their main implications are treated in the articles, the discussion and the main conclusion are presented in the introductory sections.

The first research question deals with the nature of sociotechnical systems, and how change processes in such systems should be understood. The inquiry concludes with the advice to place the emphasis on the relations of the system rather than the properties of the different components, which in this study are represented by new IO technologies. The advice is substantiated by a renouncement of a deterministic view on change processes, and an acknowledgement of the contribution to change processes from actors, regardless of their belonging to a human, technological or organisational domain.

The second research question initiates an exploration into the informal coordination of work processes in the face of contingencies. The articulation work resulting from a minor incident during a drilling operation is described in detail. This description highlights a type of informal coordination work that has a tendency to be disregarded in the more formal descriptions of work that may form the basis for the development of tools and work processes associated with IO. Additionally, the articulation work itself draws attention to distinct work processes that under smoother operations are not visible by other than those performing it. Articulation work thus has important qualities that might be unintentionally threatened in the change process of IO. The description and acknowledgement of articulation work is an attempt to reduce this threat.

The third research question addresses the interpretation of data in multidisciplinary teams by the elaboration of a case study. This case study explores a drilling operation and the role of shared understanding across disciplines with different operational goals. The main finding in this study is that in the presence of multiple goals, shared understanding is not a prerequisite for collaboration and thus the level of shared understanding is not necessarily important for the outcome of collaborative work. A consequence of this is that a focus on IO technologies that promote shared understanding through automated interpretation and shared visualisation might be unproductive and thus should be considered critically before implementation.

The fourth research question initiates a search for matters of concern associated with the implications of new IO technologies. Based on interviews with a range of IO stakeholders, this search results in a list of domains – or controversies – that are expected to be affected by new IO technologies. However, whether these domains will be affected in a way that will improve or weaken the safety of the operations is not determined by the functionality of the tools. It is argued that to control the way in which the domains are affected, they must be addressed consciously by the design, implementation and use of the technologies. This study is meant to contribute to this consciousness.

The findings of the four articles point towards the main conclusion of the thesis: By investigating sociotechnical systems and work without assuming in advance the factors or actors that make up the systems and influences their output, novel insights about how safety is affected may be obtained. An important point is that these insights may be difficult to obtain and justify by the use of generic models and conceptualisations of safety and accident genealogies that are offered by much of the existing safety research and literature. Furthermore, the consequences of IO depend on the context into which it is introduced. It is therefore decisive to identify and elaborate the contextual conditions that will affect and be affected by IO in order to understand what difference IO will make in the safety of the operations.