Control centers in large-scale events entail heterogeneous combinations of off-the-shelf and proprietary systems built into ordinary rooms, and in this respect they place themselves in an interesting contrast to more permanent control rooms with custom-made systems and a large number of operational procedures. In this article we ask how it is possible for a control center that is seemingly so “ad hoc” in nature to achieve a remarkable safety level in the face of many safety-critical incidents. We present analyses of data collected in two FIA World Rally Championships events. The results highlight three aspects of the workers' practices: (a) the practice of making use of redundancy in technologically mediated representations, (b) the practice of updating the intersubjective understanding of the incident status through verbal coordination, and (c) the practice of reacting immediately to emergency messages even without a comprehensive view of the situation, and gradually iterating one's hypothesis to correct the action. This type of collaborative setting imposes special demands to support the practices of absorbing, translating, and manipulating incoming information.

In the year 2014, the Helsinki Metro is planned to be fully automated. This automation means that the metro trains will be computer-driven and monitored remotely from a stationary control room. To investigate the challenges related to this scenario, we decided to study the ways in which the current train drivers contribute to the metro system. We conducted three separate but interrelated studies, which were based on the Core-Task Analysis method. Our results suggest that there is much more to driving the metro train than meets the eye. The drivers do not only operate the train on track and its doors at stations, but they also contribute to a variety of other important, albeit more hidden, functions in the metro system. For example, the drivers anticipate, observe, interpret, and react to events in the surrounding environment. Furthermore, they are a significant interaction link between different actors of the metro system. Our conclusion is that if the identified critical roles of the drivers are not accounted for, a migration to a fully automated metro system can affect the quality of service and raise safety issues. In addition to automated metros, the results of this research can be applicable to automation implementations also in other domains.