by Elena Psyllou
Every day we make thousands decisions; some are unconscious and others need extra effort. For instance, on my way to work, every morning I get on the Bakerloo line, change at Embankment and get off at South Kensington. But when the service suffers from severe delays, that is when I need to change my route and find the shortest and most comfortable journey to my destination.
Decision theory has gathered interest across disciplines from economics to engineering and social science. In aviation, there have been studies into the decision making of pilots. Models were then developed in an attempt to improve the design of the systems that are routinely used whilst promoting awareness across individuals of the complexity of human cognitive activity. Models like FOR-DEC and SAFE (read more here) have been used in pilot’s training and they outline the main steps of the decision making which are seek information, assess the options, make a decision, take the action.
Moving beyond such simplified versions, extensive research has been conducted in order to better understand the rationale behind the decisions and elicit the contribution of key factors such the expertise, the conditions decisions are made and the quality of information.
Research in naturalistic decision making has shown that expertise plays a key role in responding to critical situations where time is limited or information is uncertain and unavailable. Aviation and evacuation of buildings are great examples of such challenging environments in which critical decisions need to be made in a speedy manner and be of an appropriate nature.
In the 90s Gary Klein developed such a naturalistic decision making model, known as the Recognition Primed Decision (RPD) model that describes how experienced individuals make decisions in time-pressured situations where they are not able to generate and assess a range of options. The model was developed based on fire-fighters and pilots. Individuals use their expertise to find a satisfactory decision/action rather than that which is the best. They start this dynamic decision making by assessing the situation based on patterns of cues. This situational awareness activates mental models and action scripts that the individual has available from prior experience. Under such conditions, a successful decision relies on a correct assessment of the current situation, an adequate utilisation of cues, effective evaluation measures and of course, the experience of the individual.
The RPD model shows that the analytical approach of evaluating every possible course of action in order to reach the best course of action is time consuming and in safety critical industries like aviation, delays can be catastrophic. There are various ways that expertise is transformed to be easily retrieved. Research in pilot decision making (such as O'Hare et al. 2009, and Hunter et al. 2009), pilots code their previous experience into rules and cases and they retrieve those during their decision making. Rules are conditions that need to be met, for example, select a route that is straight, direct and above the minimum safe altitude. Cases are representations of events that they have previously experienced, for instance, the weather changes fast along a certain route and a diversion is needed.
Traditionally, pilots use paper charts to select the route and do the calculations for weather adjustments and fuel. Today, apps on our smartphones and tablets tell them the fastest and most direct route and with limited mental effort needed by the pilot. Similarly, route planner apps for urban travel such as CityMapper and Google maps suggest routes based on a series of computations by their algorithms. The apps not only show the flight path but they also present the trip duration and other relevant information. Nevertheless, the user who can be a pilot, a pedestrian, a motorist, remains the person that makes the decision which route to take and at what time.
Research conducted with recreational pilots and the use of route planning apps for air travel in the UK, Norway and Finland at Imperial College revealed four types of uses of these apps:
1. “YES MAN” The pilot accepts the suggested straight direct route and will make modifications in-flight, if needed, i.e. tactic decision making. A previous plan saved on the app might be used.
2. “HOLD ON A MINUTE” The pilot inspects the route and make any adjustments that will make the flight more pleasant, more cost effective and safer using rules and cases pre-flight
3. “ASSIST” The pilot uses the app to seek information, make a few calculations, e.g. regarding the weather, or see previous flight routes that are saved on the app
4. “TRUST ME” The pilot does not use the app and relies on his/her ability to remember and seek support from other sources.
It was further revealed that the pilots use the app differently based on the features of the route. In particular, for routes that are frequently used, pilots do not necessarily study every parameter (YES MAN or ASSIST) whilst routes that are flown for the first time, more time will be spent to review the airspace and weather (HOLD ON A MINUTE). Individual factors also affect the decision making and the use of the apps. In particular, pilots who want to encounter the latest weather update will do further modifications in-flight instead of pre-flight. Such en-route decision can also lead to issues such as loss of situation awareness in that pilots fail to holistically comprehend the situation in such dynamic conditions aviation operations and can lead to loss of situation awareness.
Similarities on how the route is selected using the apps are expected between pilots and any type of commuter. Sometimes we rely on the app especially if we are strangers in the area and other times we select routes we drove before. Next time you use the route planner app, as a pedestrian, a driver or pilot, think of how they contribute to your decision on what direction to follow and identify yourself which of the four types of users you are.