THE ILLUSION OF ATTENTION: Are There “Gorillas” in Your Plant?
By Wasileski, Robert F | |
Proquest LLC |
Focusing too much of our attention in a particular area can cause us to miss the obvious. Simply being aware of the phenomenon of inattentional blindness is an important first step toward reducing its effects.
The ability to recognize hazardous conditions is essential to managing risk. However, considerable research in the cognitive sciences has shown that our ability to observe our surroundings is limited by attentional resources. The failure to notice an unexpected stimulus in your field of vision while performing other attention-demanding tasks is a cognitive phenomenon known as inattentional blindness (1).
The best-known study demonstrating inattentional blindness is the "invisible gorilla" experiment (2), which asked participants to watch a video of people passing a basketball and count the number of times the basketball is passed to each participant. In the video, a person dressed in a gorilla costume walks through the scene where the people are passing the ball (Figure 1 ). After watching the video, participants are asked if they saw anything out of the ordinary. About half of them did not report seeing the gorilla.
The invisible gorilla experiment illustrates the failure in our cognitive ability to notice an unexpected stimulus in our field of vision. The illusion of attention is even more pervasive as it extends to memory and the gaps between what we think we remember and what we actually do remember.
This article presents evidence obtained through an established process-hazard-identification program that inattentional blindness is a near and present danger. This evidence is consistent with the results of the invisible gorilla experiment, and suggests that our ability to recognize hazards can be impaired more than we think. The article then provides a proactive model for organizational manage ment of change (OMOC) that addresses this phenomenon by seeking an optimal level of awareness. Recognizing that awareness is largely a function of training, knowledge, competency, and culture, the model balances experiential knowledge against the detrimental effects of emergent inattentional blindness. This approach can enable improvements in conduct of operations and operational discipline.
The illusion of attention
Most of us have experienced a situation like this: You're driving to your next engagement after a particularly challenging day at work and your thoughts begin to wander. An employee at your manufacturing site has been injured due to an unrecognized hazard. The safety manager asked you to participate in the incident investigation, document the contributing factors, and share the learnings with the organization. You reflect on how, over the past two years, company safety performance metrics have steadily improved toward the goal of zero injuries, with very few incidents during this period - and you are disappointed about the broken streak without a lost-time injury.
As you are driving, your mind has wandered - contemplating whether this incident was due to poor decision-making on the part of the employee or a cascade of highly unlikely circumstances. In the course of this reflection, your attention has shifted from driving a two-ton motor vehicle at 60 miles per hour, to recollecting incident statistics and brainstorming solutions to prevent future incidents. The vehicle has traveled more than a mile, passing several exits and numerous other vehicles. Your attention on the road has been hijacked by your thoughts.
This scenario can be understood in terms of the phenomenon of inattentional blindness. Any task that places a high demand on our attention can take away from our ability to recognize rare, unexpected, and potentially hazardous or catastrophic situations. Consider trying to multitask while listening in on a conference call and working on the computer. We cannot apply the same level of attention to both tasks because the mind's capacity to execute these tasks is limited. Why should concentrating intently on a specific task and driving be any different? It's not, and cognitive scientists have reproduced these effects in controlled laboratory experiments (3). In fact, the illusion of attention extends beyond visual perception; it's observed in hearing, memory, and even knowledge (4).
The existence of these limitations on our ability to process information about our surroundings has important implications on process safety programs in the chemical process industries (CPI). Many companies have implemented process safety and risk-management programs to comply with regulations and reduce risk to stakeholders. These include a broad range of programs that address facilities, technology, and people. Hazard-identification and risk-assessment programs frequently strive to develop employee competency in order to increase knowledge and awareness of hazards in the workplace. While important, these programs often do not explicitly address the cognitive aspects of operational discipline. Given that injuries and incidents with catastrophic consequences continue to occur, it is apparent that a gap exists in our current understanding of the importance that cognitive science plays in conduct of operations and operational discipline.
Although it might seem counterintuitive, highly focused attention may detract from our ability to see the unexpected. Focusing too much of our attentional resources in a particular area can lead us to miss the obvious.
Inattentional blindness
Research performed at
In order to see, one must look; but merely looking is not sufficient for seeing. This is apparent in motorcycle safety statistics. Intuitively, it would seem that reflective and brightly colored clothing would help make the motorcyclist stand out. However, many motorcyclists are not seen precisely because, even though they do stand out, they are unexpected. A study conducted in several
The implications for process safety are significant. In any organization, invisible "gorillas" lurking in our midst have the potential to cause damage. Whether lingering in an operating unit, in process safety documentation, or in routine operating and maintenance tasks, the tendency to not recognize obvious warning signs is prevalent.
Focusing on a specific objective takes away from our ability to notice the unexpected. We reliably observe the onset of things like cost overruns, dissatisfied customers, and production shortages, and we react. However, signs leading up to catastrophic outcomes are more subtle. Furthermore, employees tend to not recognize or give credibility to the patterns because in many cases they have not experienced these events firsthand.
Understanding the nature of inattentional blindness can explain why warning signs are not always recognized. In other words, failing to acknowledge a warning sign may not be due to conscious decision-making (seeing and not reacting), but rather due to an unconscious lack of awareness (looking and not seeing). The lack of awareness and an organization's inability to collectively act on the warning signs contribute to catastrophic outcomes. The organization is constrained in a state where process hazards, safety system impairments, and unrevealed failures go unrecognized.
Unfortunately, there is little evidence that differences in people's attention capacity or expertise influence their ability to recognize the unexpected. In fact, inattentional blindness is not always a problem; it is a result of our unique ability to focus our mind. Despite the challenges it poses, simply being aware of the phenomenon is an important first step toward reducing its effects. Furthermore, an organizational culture that is open to unconventional methods of identifying and managing hazards can enable employees to recognize patterns that would otherwise go unnoticed.
It is with this mindset that we undertook an investigation to determine the effects of inattentional blindness in order to increase employee knowledge and organizational support for process safety.
Spot the hazard
"Spot the Hazard" is a monthly contest that challenges employees to recognize process hazards. A photograph of an actual hazardous situation is distributed electronically to all plant personnel, and the members of the process safety department review submissions to identify correct responses. This program is an effective educational tool that increases employees' hazard awareness, and supplements other programs and initiatives to foster process safety culture in an organization (8).
For one investigation, we obtained a photo from an external source that showed a decommissioned storage tank with severe degradation to its insulation and structural supports. The most unusual hazard associated with the picture was that a grizzly bear had accessed the tank farm and made its way to the top of the vessel. The bear was conspicuously perched on the roof, stooped over near the relief device.
Akin to the
Management of change (MOC)
Effective, proactive organizational management of change can increase the ability of employees to recognize unexpected hazards. In order to maintain operational discipline, change must be managed at a rate appropriate to effectively maintain knowledge, commitment, and awareness. Too much change in an organization can cause knowledge attrition and inefficiency. Too little change can result in normalization of deviance (NOD) and complacency.
Consider an organization that has not experienced a significant incident within the collective memory of its workers. This lack of incident history can create a culture whereby employees fall into the trap of doing things the way they have always been done. Furthermore, employees likely will not possess the desired sense of vulnerability that is otherwise instilled in individuals who have experienced irreversible consequences firsthand. Yet, over time, as the manufacturing process undergoes subtle changes in its operating and maintenance practices, the facility may experience risk creep and a gradual rise in change blindness. This increasing risk can go completely unnoticed.
CCPS's book Guidelines for Management of Change for Process Safety (9) identifies the key principles and essential features of MOC systems. One of these essential features is the identification of potential change situations. In order to evaluate the risk of a change, the potential change itself must first be recognized. MOC workflow diagrams typically recognize this with "identify the need for change" as the first step of the MOC process.
This implies that MOC is generally prompted in response to another action. Examples of actions that may give rise to the need for MOC are capital project proposals, remedial responses to process-related incidents, and newly promulgated codes and regulations that require a facility modification. As a result, MOC has traditionally been regarded as reactive in nature. However, the potentially detrimental effects of inattentional blindness make it important for process safety management (PSM) programs, in particular MOC programs, to proactively address this issue.
Cross-pollination
Employee complacency and normalization of deviance are potential risks to an organization when too little change occurs. Placed in the context of roles and responsibilities, complacency may develop and grow in an organization that allows departmental silos to be created. Silos will almost certainly create artificial barriers at the unit, plant, and corporate levels that inhibit the exchange of ideas and the sharing of knowledge that are broadly beneficial to process safety management programs. In recognition of this, it is common practice to use an outside third party to facilitate process hazard analyses (PHAs) as a means of combating inattentional blindness.
Similarly, the practice of using employees from other units or plants as third parties in a PH A study, commonly referred to as cross-pollination, can be an equally effective strategy. Cross-pollination can also be applied to other traditional process safety activities, such as compliance audits, incident investigations, and pre-startup safety reviews. When this is done in a purposeful manner, the organization benefits from the perspectives and knowledge introduced by the diverse group of participants.
Consider the pre-startup safety reviews (PSSRs) that are conducted near the end of construction activities prior to startup (which for a large project with a construction phase of a year or more may take several weeks). Given the importance of the PSSR to the operational safety of the project, it is prudent for organizations to explicitly address team diversity in their corporate engineering standards, for example, with a statement such as: "Commensurate with the scope, novelty and complexity of the project, the audit team shall include personnel that were not directly involved with the project" (10). In this way, cross-pollination becomes a mandatory expectation of PSSRs.
Facility inspections, also known as planned general inspections (PGIs), deserve special mention because they are a vital component of a comprehensive PSM program. PGIs are conducted in an organized, planned, and recurring fashion. They are, in short, process hazard reviews taken to the field, whereby plant personnel spend 1-2 hours surveying a portion of an operating unit for both process and occupational hazards. A PGI is an excellent opportunity to identify gorillas in a plant.
Many factors can significantly hinder the effectiveness of PGIs, and PGIs that are not conducted properly may fail to address less-obvious concerns (11). Plant personnel are inclined to miss the gorillas in their own plant. Therefore, PGIs should be conducted by cross-pollinated, multidisciplinary teams. Through cross-pollination, the hazards that have been previously overlooked through inattentional blindness are more likely to be identified and addressed.
A new model for proactive organizational management of change (OMOC)
We propose that personnel awareness with respect to hazard identification can be illustrated by the Weibull probability-distribution function and the U-shaped bathtub curve (Figure 2). The nuclear power industry has used human cognitive reliability models based on the Weibull distribution to model task execution by plant operators in accident scenarios (12). The bathtub curve has been used to illustrate the concept of error states in human systems and the progression of an accident through phases as a function of increasing employee experience (13). Thus, we can use the time-variant model illustrated in Figure 2 to represent organizational hazard awareness.
In the current model, hazards that are visible to the individual are considered opportunities to identify a hazard; in other words, they are demands placed upon the individual's awareness. The average rate of failure to observe a given hazardous condition is represented by the failure rate (lambda, X) specified in the Weibull distribution function (Figure 3).
In Phase I (i.e., the learning curve phase), as an employee is indoctrinated into a new role or organization, he or she must navigate a learning curve, acquiring the knowledge and skills needed to effectively recognize unexpected hazards. Over time, as the individual acclimates to the work conditions, processes, and culture, the observation failure rate declines. In Phase II (i.e., the competent phase), the observation failure rate is lowest, and optimal hazard awareness is achieved. In Phase III (i.e., the complacency and normalization of deviance phase), abnormal and hazardous conditions are more readily accepted by employees, leading to an increased failure rate and a decline in the employee's ability to recognize hazards. Thus, personnel awareness can be characterized by three phases in terms of an individual's ability to recognize hazards (Figure 3):
* learning (decreasing failure rate)
* competence (low failure rate)
* complacency and normalization of deviance (increasing failure rate).
If OMOC is implemented proactively, continuing education and synergies among plants and units can shorten the duration of the learning phase. Similarly, the complacency phase can be diminished in magnitude by changing employees' roles and responsibilities in a deliberate manner. Thus, through proactive OMOC, an organization can influence the shape of the bathtub curve, for instance, by transitioning individuals through positions at time intervals that achieve a minimum average failure rate across the entire organization.
The minimum failure rate for the overall population is denoted XMw, and the optimal time interval to maintain a minimal rate of failure is introduced in the current model as the Wolf-Wasileski interval parameter, t^. This optimal time interval occurs in the minimum region of the bathtub curve, which is actually composed of two separate curves for early-life and wear-out behavior. The early-life and wear-out curves may be asymmetric, and the minimum (and tw) may not occur at the intersection of the two individual curves.
In Figure 3, xww is shown as a range that spans the minimum value of the bathtub curve. It is important to recognize that the optimal time interval is likely to be different depending on the characteristics of the organization and the individual unit operation. Unit operations that are highly complex and inherently high-risk typically require a longer time to enter the tmw region than a simple low-risk operation. Consequently, a large-scale petroleum refining operation would be expected to exhibit a xww time interval parameter that is shifted to the right on Figure 3, whereas the competency phase for a simple and inherently low-risk operation, such as a packaging operation or in a warehouse facility, would be shifted further to the left.
This conceptual framework suggests that an organization can achieve an optimal level of awareness by managing the rate of organizational change (e.g., Wolf-Wasileski interval parameter) at the organization, facility, and/or process unit levels. The optimal interval depends on an organization's size and dynamics, so a universal interval is not recommended. However, companies should establish a specific target interval through proactive OMOC to maintain optimal hazard awareness.
Closing thoughts
An organization's management must recognize that inattentional blindness is a near and present danger in the CPI, and that even subject-matter experts operating in their field of expertise are vulnerable to it.
In practice, the concepts discussed here can be institutionalized through corporate policies, standards, and procedures. In this way, OMOC becomes a proactive, rather than reactive, activity. Extended further, proactive OMOC can include rotating certain roles or responsibilities on a predetermined frequency. For example, in petrochemical complexes, it is common for engineers to serve threeto fiveyear terms on one process unit, and then rotate to another process unit within the complex for a second term, and so on. While this practice clearly expands the engineer's experience, it also provides a greater benefit to the organization at large because it fosters a culture of knowledge exchange and continuous improvement, which generally improve the quality of inspections, reviews, and other PSM activities. It is, therefore, imperative that organizations routinely practice some form of role rotation, particularly at the plant level.
We are currently investigating a methodology that allows facilities to quantify the Wolf-Wasileski interval parameter, tmw, for specific situations.
This article is based on a paper presented at the
An organizational culture that is open to unconventional methods of identifying and managing hazards can enable employees to recognize patterns that would otherwise go unnoticed.
In order to evaluate the risk of a change, the potential change itself must first be recognized.
LITERATURE CITED
1. Mack, A., and I. Rock, "Inattentional Blindness,"
2. Simons, D. JL, and
3. Memmert, D., "The Effects of
4. Simons, D. J., and
5. Trafton, D., et aL, "The Invisible Gorilla Strikes Again: Sustained Inattentional Blindness in Expert Observers," Psychological Science, 24 (9), pp. 1848-1853 (July 17,2013).
6. Jacobsen, P. L., "Safety in Numbers: More Walkers and Bicyclists,
7.
8. Wasileski, R. F., "Spot the Hazard! A Cultural Extension of Hazard Identification Training," Process Safety Progress, 28 (2), pp. 200-206 (June 2009).
9.
10. Wasileski, R. F., "Pre-Startup Audits," NOVA Chemicals Loss Prevention Standard 9.7, Rev. No. 5B (Oct. 2013).
11. Phillips, K. G., et aL, "Hazard Awareness Training in Managing Operational Risks," 51st
12. Apostóla Ids, G. E., et aL, "A Critique of Recent Models for Human Error Rate Assessment," Reliability Engineering and System Safety, 22 (1-4), pp. 201-217 (1988).
13. Duffey, R. B., and
ADDITIONAL RESOURCES
Gavrilov, L-, and
Nara, L., "Normalization of Deviation - Identify It Correct It Prevent It" Chem. Eng. Progress. 110 (5), p. 25 (May 2014).
Copyright: | (c) 2014 American Institute of Chemical Engineers |
Wordcount: | 4058 |
Black business promoters to celebrate 25th year
Advisor News
Annuity News
Health/Employee Benefits News
Life Insurance News