Safety

New Frontiers Identified in Human-Factors Research

Stressed-performance testing and performance indicators are among potential subjects identified as areas of potential interest in future human-factors research.

hsenow-2017-06-science-of-human-factors-hero-300x169.jpg

In response to incidents such as the explosion of the Deepwater Horizon in April 2010, the oil and gas industry has worked to generate methods that help ensure safe and environmentally responsible offshore operations. Despite these efforts, a research fellow at the Ocean Energy Safety Institute (OESI) argued that incident prevention methods will not be effective unless industry generates facility, equipment, and system designs that consider potential human-factors issues.
At a joint forum titled “Human Factors To Support Safer and Effective Offshore Energy Operations,” held by OESI and the Human Factors and Ergonomics Society, S. Camille Peres spoke about the progress being made in researching the effects of human factors in offshore projects. Peres is an assistant professor of educational and occupational health at the Texas A&M University School of Public Health.

In her presentation, Peres discussed the role human factors can play in major incidents, focusing primarily on the issues surrounding the Deepwater Horizon explosion. She said that incident and subsequent incidents around the world, such as the leaking of 6,000 tonnes of gas from Total’s Elgin platform in the UK North Sea in March 2012 and the November 2012 explosion of a production platform owned by Black Elk Energy in the US Gulf of Mexico (GOM), have made the subject even more important for owners, operators, and their onsite workers.

“We’re still seeing these incidents happen,” Peres said. “It’s not like Macondo is the be-all/end-all with regards to what we need to be focusing on. And so, it’s really time for us to take a serious reflection and talk together as a group from different industries, from different focuses, to be able to integrate what we know is impacting these incidents.”

In December, OESI released a paper that summarized the existing academic literature devoted to the adoption and integration of human factor methods, principles, and processes. [Read the paper here (PDF).] Peres said one of the major findings that came out of its study was that much of the peer-reviewed science on the subject focuses on the UK North Sea, with several studies based on the explosion of the Piper Alpha platform in July 1988. While there is some research originating from the US, she said there is not much that specifically deals with the GOM. In addition, she said there is a lack of research demonstrating the effectiveness of specific solutions related to human-factors-related issues.

“There are a lot of groups that are generating solutions, things that will help with safety climate, situational awareness, and things like this; but, right now, unfortunately, we don’t have much available with regard to the efficacy of these solutions or how much they changed incident numbers,” Peres said. “Did they improve performance in a way that we wanted to see? We’re not seeing this in the peer-reviewed science, this evidence that will let us know what the industry needs to be spending its money on. For practitioners, you want to be able to go to your bosses and say, ‘I’m asking you to write this big check for career resource management, and here’s the evidence that it works.’”

In the paper, Peres et al. argued that the lack of actionable research was in part because of the ethical difficulties involved in setting up experiments that could accurately reflect operational conditions: An experiment that affords attribution of cause and effect in a hazardous condition increases the chance of a disaster that puts people at risk. With that in mind, an operator looking to make decisions on such things as rig design, safety programs, and training in research must often take inferential leaps.

Despite this handicap, the paper suggested ethical methods of conducting empirical studies to identify casual relationships between the variables that often affect human factor performance. One such suggestion was for researchers and businesses to use pre/post designs that compare the outcomes of interest—such as the number of safety incidents and the effectiveness of performance—before and after various intervention techniques like interface redesign and the implementation of an alarm management strategy. Scenario-specific training facilities can also be useful for operators.

In addition to identifying gaps in human-factors research, Peres et al. examined potentially important areas of future human-factors research:

  • Perceptual-based vs. cognitive-based decision making. This involves studying the level of human information processing required to make a decision once all information is available. Cognitive engineering emphasizes the design and development of complex systems while taking into account human behavior. While not a new concept, Peres et al. argued that one threat of research yet to be mined is the identification of user interface components such as readouts, figures, and graphs that allow workers to make decisions based on lower-order, fundamental human information processing capabilities instead of demanding more cognitively expensive capabilities.
  • Instantiating “super workers” wisdom—The paper suggests that companies may be able to incorporate the knowledge of their more highly skilled workers (the “super workers”) into rig control software and facility operation. Peres et al. argue that more research must be performed on identifying the thought processes that allow these workers to identify trends quickly, presenting this information in a way that would allow a new or mediocre worker to improve his or her performance, and ultimately building the information into artificial intelligence systems that mitigate or remove the element of human error.
  • Vigilance—As automation assumes a greater share of onsite operations, the industry must look into designing systems that help workers maintain attention and focus and whether there may be value in insinuating fake error conditions at nonroutine intervals.
  • Stressed-performance testing—The use of environments and facilities that simulate life-threatening conditions to see how people perform under stress. These simulation environments are typically designed for emergency-response training. The paper lists the Texas A&M Engineering Extension Services Fire School and Disaster City as an example.
  • Return on investment—This involves addressing the business case for human-factors research and developing cost-effective methods for retrofitting existing facilities. Peres et al. argued that more research must be done to identify which human-factors evaluation methods, applied at which specific points in the design of a drilling system, yield the best returns.
  • Key performance indicators—KPIs may also be valuable in helping companies know whether their systems and programs are making meaningful progress toward successfully integrating human factors and ergonomics. Peres et al. write that one of the main challenges in developing KPIs is the high amount of effort needed to generate an indicator with sufficient specificity for measurement.