Modeling and simulation of offshore workers' behavior

Musharraf, Mashrura (2018) Modeling and simulation of offshore workers' behavior. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[img] [English] PDF - Accepted Version
Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.

Download (4MB)


The offshore oil and gas industry functions in a team work culture in which operations depend not only on individuals’ competency, but also on team skills. Team skills are even more necessary when it comes to handling emergency conditions. Emergency conditions are dynamic in nature and personnel on board are challenged with evolving high-risk situations, time pressure, and uncertainty. One way to effectively handle emergencies is to train personnel to a competency level, both individually and as a part of a team. This would increase the chance of achieving safety in a timely manner using the available resources such as information, equipment, and people. Such training involves enhancing team members' understanding of human performance, in particular, the social and cognitive aspects of effective teamwork and good decision making. Post-accident analysis of offshore accidents shows that conventional training programs are often too generic, and that they are not designed to identify and tackle the human factors that are critical for evolving offshore emergency situations. Recognition of the importance of human factors on operator performance raises the need for training that goes beyond conventional training programs and incorporates non-technical training focusing on leadership, command, decision making, communication, and teamwork. A major difficulty to design such training is that it involves practicing emergency exercises with a potentially large number of participants, each playing the appropriate role in a given scenario. Such large-scale team exercises suffer from both organizational and educational drawbacks. The amount of human and financial resources needed for such a training exercise is difficult to organize. Furthermore, it is very hard, if not impossible, to get all team members together at the same time and location. Also, the team members may have variability in the competency levels (novice versus advanced trainees) and hence different training needs. One effective and flexible solution to this problem is to use intelligent artificial agents, or ‘virtual workers’, in a virtual environment (VE) to play different roles in the team. Virtual workers are artificially intelligent agents that can reproduce behaviors that are similar to or compatible with those of a real worker. This research proposes to develop a human behavior simulation model (HBM) that can be used to create such virtual workers in the context of offshore emergency egress. The goal of this research is to develop a human behavior model that can simulate offshore workers’ emergency response under the influence of performance influencing factors (PIFs). The first part of the work focuses on understanding human behavior during offshore emergency situations. A two level, three factor experiment was conducted in a virtual environment (VE) to investigate the relationships between the PIFs and human behavior. Influence of both internal and external PIFs were investigated. Knowledge acquisition and inference processes of individuals were also investigated in the experimental study. In the second part, a computational model was developed to capture the across-subject variability observed during the experiment. Interviews with subject matter experts (SME) were conducted at this step to ensure that the model is able to produce a realistic range of human behaviors. The final step was to validate the developed behavior model. All high-level tasks to validate the HBM were performed. Special emphasis was given on acceptability criteria testing to ensure that the integrated HBM performs adequately under different operating conditions.

Item Type: Thesis (Doctoral (PhD))
Item ID: 13225
Additional Information: Includes bibliographical references.
Keywords: Bayesian network, Decision tree, Offshore emergency egress, Human behavior model, Virtual operators for simulators
Department(s): Engineering and Applied Science, Faculty of
Date: March 2018
Date Type: Submission
Library of Congress Subject Heading: Collective behaviour -- Mathematical models; Offshore oil industry -- Accidents -- Mathematical models; Emergency management

Actions (login required)

View Item View Item


Downloads per month over the past year

View more statistics