Naturalistic Decision-Making In Aviation: Strategies And Models

Cognitive Levels of Errors

Rasmussen (1983) developed a conceptual model, which can be applied to the design of the man-machine interface, and has the potential to reduce the potential for accidents. The performance of human brain can be characterized as goal-oriented or rule-controlled. Based on cognitive aspects of human behavior, there are three associated errors, namely, skill-based, rule-based, and knowledge-based error. This classification is commonly known as SKR (skills, rules, and knowledge) scheme. This theory states that the human operator is at one of the three cognitive level, which depends on the nature of the task and the associated level of experience.

The skill-based performance error is the error of execution, and this type of error is made by extremely experienced people (Cacciabue, 2013). These people do not have to align different pieces of information in the information interpretation and all the information processing is done at the sub-conscious level. The reaction to the cues is automatic like walking or riding a bicycle. The responses are based on pure stimulus-level, which are developed at neurological level, and performance is governed at an automatic level. Rule-based performance error is advanced error, which is faced when there is inability to identify the situation or circumstances. These errors are due to both experience and training. The individuals who commit this type of error are familiar with the task; however, do not have the amount of experience to accomplish it at subconscious level. The cues are identifies and recognized, and based on the previous experience appropriate actions are taken.

The highest level of error is knowledge-based (Mital, Kilbom & Kumar, 2000). In these types of error, people do not apply the skills and rules to the task or situation, but they apply the previously learnt information or the previous experience to the new task (Kletz, 2008). The individuals working at the knowledge-based level have challenges in grasping knowledge, or limitation in applying the existing knowledge to the current situations.

The naturalistic decision-making is based on the fundamental of bounded rationality. It states that rational decision-making can only yield better decisions when there is time for taking a decision, the goals and the issues are well defined for the organization or the decision-maker (Podofillini et al. 2015).

It can be critiqued that most of the decisions and incidents in aviation are the result of poor decision-making and judgmental strategies. The pilots are not taught to take decisions in highly dynamic and risky environment. The traditional decision-making models are unable to explain the decision taken by the pilots in complex accidents. It is because the pilots do not have ample time to identify the root cause and evaluate different strategic choices.

Decision-Making Models in Aviation

The classical decision-making model states that the managers should always be logical and rational while taking decisions. However, it is not possible in the dynamic environment of aviation, as the pilots have to take quick decisions at a very small duration of time (Zsambok & Klein, 2014). The laboratory decision-making model is focused on identifying the optimal solution to the problem. It assists the managers in identifying the best decision for a specific problem. This model helps the decision-maker in problem structuring, determining the criteria for identifying the solution and recognizing the best solution for a particular problem. However, it is a time-consuming process; therefore, it cannot be applied to aviation.

 The naturalistic decision-making is an intuitive strategy, which can be used by the pilots to make operational decisions. The cockpit of airplane can be categorized as naturalistic environment due to several characteristics such as experienced pilots, different team players, dynamic operational conditions, dynamic goals, high risk, minor time-window, and ambiguous or unavailable data. NDM strategies can be considered as an intuitive decision strategy used by the pilots to make intuitive decisions. The strategies can assess the situation; evaluate the decisions through mental evaluation. The intuitive and analytical decision-making strategies are important in several situations. It includes situations in which there are limited goals, limited time constraints, and correct set of information (You, Lu & Chen, 2015). Since the naturalistic environment is dynamic, the decision maker should continuously reassess the situation and the appropriateness of situational models. The pilot should evaluate the situation and take an appropriate course of action to deal with the situation. However, the situational assessment and awareness is crucial and time consuming. The situational assessment is important for good decision-making and increasing the decision accuracy (Zsambok, & Klein, 2014).

The mental simulation is conducted before a decision action is taken. The decision maker evaluates different course of action and selects the best approach. The simulation process involves successive steps, which are required to be taken, potential outcomes of the process and how these problems can be handled. However, if complete time is not available for the mental simulation, the decision maker will implement the action, which is based on the decision. The decision will be taken based on the experience; however, subsequent changes would be made so that satisfactory outcome can be produced.

Therefore, decisions taken from other decision models are unsuitable for aviation. It is due to the reason that these methods work well only, when there is significant amount of time, information, and money. However, such situations are rarely found in aviation. There are always conflicting and incompatible goals in aviation and; therefore, a single, optimal decision cannot be made by the pilots. Therefore, optimal decision is not available and is not necessary; however, good decisions can equally serve the purpose. The result outcomes of such decisions are acceptable with less amount of resources and time being invested. The naturalistic model yields such decisions.

Recognition-Primed Decision Making Model (RPD)

With the high level of risk and little amount of time, the pilots take the first workable decision rather than comparing different decisions for optimality. There are very small chances that changing the decision from good to optimal can increase the efficiency or safety of the passengers. In different circumstances, the people choose different decision-making models, which can yield the workable solution in least resources. The focus is not on the optimal solution. However, in operational environment, the people should not invest personal resources as the decision influence several different aspect of operations of a business organization.

In naturalistic situation, achieving an optimal solution is challenging and nearly impossible. An optimal decision refers to a single, well-defined goal. However, in such cases, the situation remains stable. In naturalistic environment, goals remain dynamic and competing. In the aviation, there are several dynamic goals such as safety, profit, company policy and customer comfort. It is difficult to satisfy all of these criterions; therefore, the strategy cannot yield analytical strategy (Simpson, 2001).

The naturalistic decision-making and the operator experience are interlinked. It is an important relationship, as the naturalistic, intuitive strategies have to be taught to the learning pilots.

The recognition-primed decision making (RPD) is a model derived from naturalistic decision making in which the alternative courses of action are derived by recognizing critical information and knowledge. Different alternatives are examined; however, different options are not compared with each other. The decision makers compare the current situation with their experience and rules formed before. The decision is made by finding the similarities between different situations; therefore, the decision-making model is named as recognition-primed decision-making. There are three phases in this model, namely, situation recognition, serial option evaluation, and mental simulation (Simpson, 2001).

Since the naturalistic environment is dynamic, the decision-maker has to examine the situation and the changes in the environment rapidly. In the aviation, when the pilot understands the situation, he can take an appropriate decision. The situational analysis is not appropriate model in this situation as the awareness is crucial and it is a time consuming process. It is apt model for decision-making and can increase the accuracy of the decisions (Salas & Martin, 2017).

The last stage before taking a decision is mental simulation in which the decision action is implemented. The decision maker evaluates different courses of action by imagining the sequence of events, which will occur after the decision. The mental simulation includes the successive steps of the decision, the outcomes of these steps, problems, which might be encountered, and how these problems can be encountered. The decision maker can reject, modifies, or implement the decision. If there is no ample time for decision-making, the focus is on satisfactory decision rather than on optimal decision. In these cases, time is not available for the mental simulation, therefore, the decision-maker takes a decision as per the experience, which will be successful and provide a satisfactory outcome.

Hierarchal Task Analysis (HTA)

The hierarchal task analysis (HTA) is a broad task analysis method, which analyzes the task according to the hierarchy of the operations and plans designed according to the standard notations of the chart. The hierarchal task analysis informs the analysts when different sub-tasks should be executed so that the overall goal of the organization is met. In the task analysis, there are three level of task analysis, goals, tasks, operations, or actions. There are several benefits of hierarchical task analysis, because of which the method is popular. It is a cost-effective method, as the hierarchal description is needed to be developed when it is needed to wherever it is needed for the purpose of analysis (CCPS, 2010). The requirement of the training is negligible and it can be easily implemented within the organization. The output of the analysis is beneficial for different applications. The output provides detailed description of the task which provider great insight to the user. It is a generic method and can be applied quickly.

 This method is easy to learn and can be easily used by the people. It can serve as the basis of several different assessments. This method is commonly used and is applicable on the cognitive analysis of the customers. The HTA is powerful method to analyze different types of physical and mental activities and their application to different domains (Federal Aviation Administration, 2017).

There are also certain disadvantages of HTA such as the analysis has to develop skills to analyze the tasks and the associated techniques. The focus of this analysis is on the descriptive analysis rather than the analytical analysis. In case of complex decision-making tasks, HTA has to be used in combination with other cognitive models (CCPS, 2010). However, the output cannot be used for the design purpose. It does not meet the requirements of cognitive analysis and is not efficient for large tasks. In this method, the data collection is very time-consuming. The reliability of the method is also controversial as different analysts yield different results for the same task.

References

Cacciabue, P.C. (2013). Guide to Applying Human Factors Methods: Human Error and Accident Management in Safety-Critical Systems. Springer Science & Business Media.

CCPS. (2010). Guidelines for Preventing Human Error in Process Safety. John Wiley & Sons.

Federal Aviation Administration. (2017). Task Analysis > Hierarchical Task Analysis (HTA). Retrieved 14 January 2017 from https://www.hf.faa.gov/Workbenchtools/default.aspx?rPage=Tooldetails&subCatId=28&toolID=106 

Kletz, T. (2008). An Engineer’s View of Human Error. IChemE.

Mital, A., Kilbom, A., & Kumar, S. (2000). Ergonomics Guidelines and Problem Solving. Elsevier.

Podofillini, L., Sudret, B., Stojadinivic, B., Zio, E., & Kroger, W. (2015). Safety and Reliability of Complex Engineered Systems: ESREL. CRC Press.

Salas, E., & Martin, L. (2017). Decision-Making Under Stress: Emerging Themes and Applications. Routledge.

Simpson, P.A. (2001). Naturalistic Decision Making in Aviation Environments. DSTO Aeronautical and Maritime Research Laboratory.

You, J. X., Lu, C., & Chen, Y. Z. (2015). Evaluating health-care waste treatment technologies using a hybrid multi-criteria decision making model. Renewable and Sustainable Energy Reviews, 41, 932-942.

Zsambok, C.E., & Klein, G. (2014). Naturalistic Decision Making. Psychology Press. Liu, H. C.,