The workshop was an arena for discussing the various issues that were prioritised. The key findings and proposals for measures:
• The inadequate level of HF competence should be remedied, [T1]i.e. the need to raise the level of competence concerning HF within the petroleum sector, amongst stakeholders such as operators, clients, management, supervisory authorities and consultants/suppliers. This could be achieved through training based on research-based knowledge within HF or the use of experts.
At the same time, the HF principles/methods in the regulations need to be updated. The need for a stronger focus on HF applies from the early phase of investigations. This is especially true in connection with a higher level of automation of safety-critical operations, where technology must be adapted to human capabilities and limitations.
• Technology focus must be balanced against attention to user-centred design (particularly as regards safety-critical tasks, HMI, and non-conformity management) [T2];There is a strong impetus from the technology which must be balanced by a focus on user-centred design, with the consequence that systems must be adapted to meet user needs as soon as possible in the process, and efforts must be made to ensure that safety-critical features and non-conformities can be managed by users via thorough HF analyses.
• Insufficient use of HF standards, [T3].HF standards become increasingly important when automating more and for more complex systems which an operator has to intervene when automation fails. There are many different standards and practices, but it is important to define a set of basic standards that can be used as a basis within the industry. It is therefore important that reference continues to be made to ISO 11064, and that more reference is made to standards such as ISO 9241:210. They should be mentioned in the regulations/guidelines and followed up within the petroleum sector. This follow-up could take place in a number of arenas between stakeholders with, for example, developers, where it is important to ensure a common understanding between the supervisory authorities and stakeholders. The verification and validation of solutions based on good practice should continue, with the involvement of HF experts.
• Accident investigations have too strong a focus on technology and should cover HF/human experiences and roles, in addition to organisational factors [T5]. All accident investigations should have HF experts as part of the investigation team from the outset. If appropriate, their involvement could subsequently be reviewed following an evaluation. The investigations should describe how the humans perceived the situation, e.g. with support in taxonomy from
sensemaking or situational awareness. In order to illustrate best practice, the industry and supervisory authorities could select a number of good investigation reports which represent best practice and document the methods that were used. In this context, Accident Investigation Board Norway's report on the Helge Ingstad in 2019, or alternatively the CSB’s report on the Deepwater Horizon (US-CSB, 2016), could be used as an example.
• Investigations rarely consider design, but should consider the appropriateness of the design.[T5].Inappropriate design can be the root cause of many of the accidents, but is rarely considered by investigations. It is therefore proposed that design always be considered.
Moreover, successful designs in connection with successful barriers/features should be highlighted more often.
• Near misses should be identified and analysed to a greater extent. There are currently no defined requirements regarding what must be logged concerning safety incidents involving automated systems in the oil and gas industry, or how such data should be handled in
connection with reporting and learning. Both the authorities and the industry should jointly establish requirements regarding the data that must be logged for safety-critical automated systems which can provide information on safety-critical incidents [T4].
• Lessons from successful factors [V5]. A higher priority may be given to the sharing of experiences of successful projects, particularly as regards good work processes. There are a number of collaboration forums and networks which could be used to share learning
9 Proposals for initiatives and further work
In the following, we propose measures and further work based on the project’s findings. The findings and proposals are based on inductive analyses based on industry experience, combined with deductive analyses where we use theory from the area of human factors. The methodology used to draw up the proposals was based on the principles of action research, where practice and what the stakeholders actually do were examined (Guba et al., 1989).
A consistent finding by the project team was that there is a strong focus on technology without the necessary attention being paid to human strengths and limitations, a failing which could
compromise HSE and hamper successful automation. Automation has generally been successful where it has been introduced in well-defined areas, gradually and in interaction with users.
Automation can contribute to better utilisation of petroleum resources by enabling the drilling of challenging wells as well as drilling in formations where drilling has not previously been possible.
Automation can lead to more efficient drilling and provide support for the earlier detection of failure events through the various technical solutions used in the drilling process. The positive user- centred development of automation should be continued based on lessons learned from pilot projects. It is necessary to apply the principles of meaningful human control when automating operations, partly in order to safeguard HSE.
Figure 9.1 The human who uses a technology must be involved in communication with
authorities, in the development and design of the technology and through continuous learning and updating of the technology through operation.
In the following, we summarise findings and proposals for measures, based on the results of the project activities (where the measures [Tx] can be traced). The measures are proposed from a lifecycle perspective based on standards such as ISO 11064. The order reflects the desire for cost reductions and efficiency by ensuring that good practice is incorporated at an early stage from delimitations, analyses and design; see Figure 9.1.
Authorities and regulations
Users of new technology
Operation of automated technology
Development and design