This thesis Teams organising ‘work as done’: resilience, repetition, and expertise presents a series of studies and research papers on how teams organise their work for safety, efficiency and a host of other goals. In a sense, it aims to inform on improving how to organise teams and organise for teams, but this is expected to arise from a better understanding of ‘work-as-done’, rather than imposing a ‘work-as-imagined’. A broader view of initiatives across a number of industries is part of this thesis, however, the major empirical focus is on ‘work-as-done’. Empirically, the research for this thesis was driven by the work of semiautonomous blast crews who operate in open-pit mines and quarries, as well as electrical line crews deployed across southeast Queensland. The thesis also took a broader view that makes its findings applicable to other teams as well. The thesis is organised around four papers, which form a literature review, a methodological reflection, an empirical investigation, and a theoretical paper extending the empirical work. The thesis began by investigating the-state-of-the-art way to organising teams, by reviewing Crew Resource Management (CRM) literature across industries. The review discovered that there was very little evidence to support that CRM achieve their goals of improving safety and efficiency. In addition, there was no unified conceptualisation of how CRM is supposed to reach those goals. This diversity in conceptualisation made it impossible to further investigate whether CRM is working as intended, as the questions to investigate differed per conceptualisation. This review showed there is little support for teaching a stateof- the-art way of organising to people, nor a clear idea of how teams operate. The next study in the thesis explored the methodological considerations for investigating everyday work for safety purposes. By considering the possibilities and different schools of thought in safety, I reflected on the design of investigations of everyday work for safety. I found a general tension between trying to capture and learn from the everyday, and the goal of investigating something that is known to be relevant to safety. What is known the be relevant can be hard to fit with the reality of everyday work and could steer away from exploring new areas, while with the everyday there is the uncertainty of whether something is relevant for safety. These reflections led to the design of the methodology for the third study. In this study, I applied problem-oriented ethnography to investigate how blast crews and linesmen organise routine work. From my results, I explained how individual decisions give rise to general repeated patterns of work at a team level. I found that operators have a stable problem understanding of and repeating patterns in how they approach a task. This repetition is not a replication of actions, but in the way teams divide and structure a larger task into smaller goals ¾ their solution structure. As team members work together, their problem understanding and solution structure converges, which leads to smooth and coordinated work process of a team. The problem understanding and solution structure capture part of the expertise of crew members and allow them to perceive meaningful signs in their environment. The repeating patterns with which teams complete tasks makes them more sensitive to their operations, as deviations will stand out. In addition, crew members could interact adaptively because of it, as they could use intentional deviations to alert each other. Building on the finding that even for routine work, team members rely on expertise to be adaptive, in the fourth study I explored the relationship between expertise and resilience, and how to manage expertise on the level of frontline work, teams and management, and systems to enhance resilience. The conclusions from this thesis explain how teams organise routine work resiliently. Unlike the work on High Reliability Organisations and Resilience Engineering, my conclusions stress how repetition helps teams collaborate and cope with complexity, while still allowing for the needed adaptations. My conclusions challenge the distinction between heedful and routine action from the High Reliability Organisations literature. The thesis extends theories from Cognitive Systems Engineering on how team processes can shape the goals of individuals. My conclusions contradict organisational routine theories on what repeats in a routine and extend theories of organisation routines as to how actions in a routine link together. In addition, I made methodological contributions to investigating of everyday work for safety and the study of changing routines. In terms of practical application, the thesis makes suggestions on team member composition and rotation, as well as on the specification of procedures.