Abstracts:Patients are continually being put at risk of harm, and health care organisations are struggling to learn effectively from past experiences in order to improve the safe delivery and management of care. Learning from incidents in health care is based on the traditional safety-engineering paradigm, where safety is defined by the absence of negative events (Safety-I). In this paper we make suggestions for the policy and practice of learning from incidents in health care by offering a critique based on a Safety-II perspective. In Safety-II thinking safety is defined as an ability - to make dynamic trade-offs and to adjust performance in order to meet changing demands and to deal with disturbances and surprises. The paper argues that health care organisations might improve their ability to learn from past experience by studying not only what goes wrong (i.e. incidents), but also by considering what goes right, i.e. by learning from everyday clinical work.

Abstracts:Research has acknowledged the relevance of accident information for prevention and learning and the need for companies to develop a reporting and learning culture. Few studies have approached this issue by comparing the different learning strategies used by companies. The aim of this study is to explore how companies use accident information and to develop strategies for learning from accidents, which cover all the learning cycle phases by: (a) identifying learning patterns across company and activity sectors, (b) checking for potential differences among certified and non-certified companies. Seventeen case studies were conducted with organizations operating in different sectors in Portugal. Data was collected from extensive, semi-structured interviews with key stakeholders and an analysis of relevant documentation. All data was subjected to a descriptive analysis, followed by multivariate analysis using Multiple Correspondence Analysis (MCA). The main MCA results showed two dimensions corresponding to the technical and social learning and four patterns were found, each corresponding to different levels of learning practices ranging from the minimal practices used to a higher degree of learning, combining practices of a technical and social nature. Additionally, the results revealed that companies in the same activity sector may have very different practices, independently of OHSAS certification. The results allow us to conclude that organizations with good safety practices tend to follow the complete learning cycle. Overall, these organizations have established procedures to report accidents and to collect information on them but there are organizations that still do not maximize their means of learning from work accidents.

Abstracts:Drivers of contemporary vehicles are now able to relinquish control of the driving task to the vehicle, essentially allowing the driver to be completely hands and feet free. However, changes to legislation taking effect in 2016 will require the driver to be able to override the automated driving systems or switch them off completely. Initially this functionality is likely to be limited to certain areas, such as motorways. This creates a situation where the driver is expected to take control of the vehicle after being removed from the driving control-loop for extended periods of time, which places high demand on coordination between driver and automation. Resuming control after being removed from the control-loop have proven difficult in domains where automation is prevalent, such as aviation. Therefore the authors propose the Gricean Maxims of Successful Conversation as a means to identify, and mitigate flaws in Human-Automation-Interaction. As automated driving systems have yet to penetrate the market to a sufficient level to apply the Maxims, the authors applied the Maxims to two accidents in aviation. By applying the Maxims to the case studies from a Human-Automation-Interaction perspective, the authors were able to identify lacking feedback in different components of the pilot interface. By applying this knowledge to the driving domain, the authors argue that the Maxims could be used as a means to bridge the gulf of evaluation, by allowing the automation to act like a chatty co-driver, thereby increasing system transparency and reducing the effects of being out-of-the-loop.

Abstracts:Effective learning is essential for a safe workplace. Through learning from incidents (LFI), knowledge is applied and embedded within the work environment in ways that can prevent future incidents. In order to improve their LFI processes, such as incident reporting and analysis, or the dissemination of investigation outputs, organisations need an instrument that allows them to diagnose the quality and effectiveness of their LFI processes, making sure that LFI leads to positive safety outcomes. This paper outlines an instrument that measures the quality of LFI processes and practices: the Learning from Incident Questionnaire (LFIQ). The LFIQ identifies employees’ perceptions and experiences of LFI processes and practices. This paper describes the validation of the LFIQ instrument via a pilot study conducted at two energy companies involving 781 participants. Through factor analysis the instrument was shown to have sufficient validity to become a useful tool for industry; by gaining insight into employees’ perspectives on LFI, frontline managers and supervisors can have evidence on which to base improvements to the local work environment and prioritise areas for improvement.

Abstracts:The present research focuses on the processes of identifying remedial actions subsequent to incidents at two Swedish nuclear power plants. Data from 106 in-depth analyses were analysed together with interviews with event investigators. The results and previous research in the domain indicated a need to further develop the process for identifying remedial actions. A method was developed that focuses on process descriptions and identifications of strengths and weaknesses inherent in the process(es) in which an incident occurred. The method uses a participatory approach with actors from the relevant process(es). A case study was conducted which showed promising results. The method is discussed in terms of generalising to a more process-oriented experience feedback than usually is applied.

Abstracts:Major accidents are complex, multi-attribute events, originated from the interactions between intricate systems, cutting-edge technologies and human factors. Usually, these interactions trigger very particular accident sequences, which are hard to predict but capable of producing exacerbated societal reactions and impair communication channels among stakeholders. Thus, the purpose of this work is to convert high-dimensional accident data into a convenient graphical alternative, in order to overcome barriers to communicate risk and enable stakeholders to fully understand and learn from major accidents. This paper first discusses contemporary views and biases related to human errors in major accidents. The second part applies an artificial neural network approach to a major accident dataset, to disclose common patterns and significant features. The complex data will be then translated into 2-D maps, generating graphical interfaces which will produce further insight into the conditions leading to accidents and support a novel and comprehensive “learning from accidents” experience.

Abstracts:The safety record of the international merchant shipping industry has shown a continual improvement for a prolonged period due to its ability to learn from incidents and prevent recurrence; through training, education, technology development and regulatory change. However, the annual rate of ship losses has remained relatively unchanged in recent years. The industry has become accustomed to a safety regulatory regime based heavily on embedding lessons from incidents. That regime has served it well for more than a century but the industry is experiencing rapid change, which presents a challenge to this approach.

Abstracts:Accidents are notoriously frequent in mountain sports, but thorough understanding of the mechanisms of accidentality remains limited by the fragmentation of sources and by mostly heterogeneous methodologies. Nonetheless, the effectiveness of prevention must rely on detailed knowledge of typical circumstances and scenarios. Rooted in the statement that an accident is never induced by a single cause but rather by a dynamic combination of factors, this paper explores the opportunities offered by a systemic analysis of experience feedbacks on accidents and close calls. The study identifies risk factors for several hundred mountaineering accident and near miss reports. In order to enhance the benefit of these descriptions and to show the interaction of a broad variety of contributing factors, it introduces graphic models. This is not an attempt to compress the unique richness of each story, but rather to create a tree structure using the concatenation of multiple testimonials, thus enabling researchers to build general lessons out of individual cases.

Abstracts:The perceptual cycle model (PCM) offers a process-orientated approach to understanding decision making. This approach is distributed in nature as it considers how internally held schemata and external environmental information interact to produce actions and behaviour. This an essential component of any incident reporting system, although it is often lacking. In its current form the PCM only provides a very high-level of explanation. This research utilised data from critical decision making interviews to deconstruct the three high-level categories of the PCM into a 28 item taxonomy. In doing so, we were able to provide a more detailed description of aeronautical critical decision making (ACDM) by demonstrating the relevance of different concepts in different phases of dealing with an incident. The data were used to model the ACDM process. The taxonomy can be used for gaining a comprehensive understanding of both the contextual and cognitive components of decision making, something that existing taxonomies and models often fail to do.

Abstracts:This paper outlines a research and development agenda for the nascent field of Learning from Incidents (LFI). Effective, deep and lasting learning from incidents is critical for the safety of employees, the general public and environmental protection. The paper is an output of an international seminar series ‘Interdisciplinary Perspectives on Learning from Incidents’ funded by the UK Economic and Social Research Council (ESRC) in 2013–2016 http://lfiseminars.ning.com/. The seminar series brought together academics, practitioners and policymakers from a range of disciplines and sectors to advance the theory, methodology, organisational practice and policy in LFI. Drawing on a range of disciplinary and sectoral perspectives, as well as on input from practitioners and policymakers, this paper lays out four key research and development challenges: defining LFI; measuring LFI; levels and factors of LFI; and strengthening research-practice nexus in LFI.