Zandi talks Aeroplane SafetyDate Released: Mon, 5 October 2015 11:49 +0200
This paper represents the primary focus of Zandi's MSc research, which will examine the alertness of pilots whilst autopilot is engaged. As such, it will investigate how to improve this to ensure a rapid transition from" passive" to "active" control of the aircraft should automation fail. Unfortunately, whilst the pilot is monitoring the system, the skills required to operate the flight system effectively and safely would likely be degraded from non-use.
Zandi clarifies her research further, by explaining that the “attentional demands” on the human operator during interaction with semi-automated operations of the flight are also in question. As has been documented by past studies such as Billings (1991), human operators working with automation have a diminishing ability to subsequently perform tasks manually in the face of automation failures . Given this background, the relevant research questions are:
- What alertness strategies could the human operator use to maintain alertness during passive supervision of the flight system?
- Whilst motor performance may remain unchanged, what happens in the cognitive workload and performance with regards to alertness with the increase in mental demands?
- What strategies could be put in place to allow for a rapid transition from passive to active participation by the human operator
The theme of the conference was “doing safety, not only talking safety”. The majority of attendees and presenters were from the industry, and Zandi reports that it was fascinating hearing the views of pilots and air traffic controllers. She has found that with their insight into the pressures of working in the industry, she is able to view her own research in context, and with a different perspective.
Zandi says, "finding that common ground between science and the ‘real-life’ operation of an aircraft enabled us to find the cracks in literature that require attention, which could then be applied in the reality of the industry".
"I gained a lot of insight from actual pilots who have been operating in this industry for years. After my presentation, a few of the pilots were very interested in this research, stating the relevance of it in the actual field. One pilot mentioned that in a long haul flight, whilst monitoring the flight system after a period of 4 hours (with monotony and fatigue setting in), the system might fail in a split-second and as the pilot, you have to perform what they term the" ANC procedure", which is Aviate-Navigate-Communicate. This allows you to prioritize what you need to do in order to get the flight back to operating safely in the chance that the automation should fail. So that for me created a platform to network and get as many contacts as I could in order to help me with my research."
Zandi's experience has highlighted the gap between industry and science, particularly in this field. She believes that something she took away from the conference is the dire need for scientists to partner with industry, to maximise the relevance of the science.
Zandi's full abstract is presented below.
The alertness of pilots during semi-autonomated flights
Automation in the aviation industry is generally acknowledged as a useful tool in reducing pilot workload, particularly during short and long-haul flight operations (Hoh, Smith & Hinton, 1987 and Beringer & Harris Jr., 1999). The handling of the aircraft is thus often shared between the pilot and flight system. The varying amounts of authority between the two present new challenges where the pilot is required to have an accurate picture of the current and future status of the flight at all time, whilst not being in full control of the aircraft (Hughes, Rice, Trafimow, & Clayton, 2009). The assistance from the automated system thus enables pilots to operate the aircraft effectively (Wickens, Helleberg, & Xu, 2002).
There is however growing apprehension about cockpit automation and its potential effects on pilot performance and alertness during automated flying, with the engagement of autopilot (Sarter & Woods, 2009). It is essential that pilots are capable of diagnosing troublesome situations in cases of different scenarios. Such scenarios include getting an overview of complex situations or simply taking over from the system, should a manual intervention be needed, where the transition needs to be of immediate reaction.
Typically, pilots are inclined to have short periods of ‘in-seat rests’ that allow them to recover from the workload of the tasks. Such tasks include flight planning, navigation, performance management and flight-progress monitoring (Sarter & Woods, 1992). Monitoring, being passive, can become monotonous and mind-numbing. This could further result in a lack of attention from the pilot or a set of skills necessary for aircraft control be degraded from non-use during the process of supervision. The anticipated research is thus aimed at the roles of pilots’ during semi-automated flying, investigating which attention strategies could aid in alertness management of the pilot, should the resumption of control be required due to automation failure. Results will be presented on at a later stage.