How my planned 30 hour journey became an 88 hour odyssey, and what that taught me about the impact of complexity on planners and their plans.
This article was originally published on Medium. The full article can be accessed at this link: https://medium.com/@aaronpjackson/your-flight-is-delayed-the-impact-of-complexity-upon-a-simple-plan-a2f70d31281
I recently commenced an appointment as Distinguished Visiting Professor at Canadian Forces College in Toronto, Canada. Before commencing this appointment I had to travel to Toronto from my home town of Adelaide, Australia. That’s literally the other side of the world and yet, thanks to air travel, getting between these two very distant locations required only a simple plan. Three flights. Adelaide to Sydney. Sydney to Dallas (via Los Angeles). Dallas to Toronto.
Picture 1: My planned route (shown in red). Source: author
Per my booking, I expected to be travelling for 30 hours. I left my house in Adelaide at 0700 on a Saturday, expecting (thanks to time zone changes) to arrive in Toronto at 2230 on the same day. Instead, after 88 hours I finally arrived in Toronto at 0850 on Tuesday morning — almost two and a half days later than intended. This article explains the cause of these delays in the context of air travel as a complex system, then offers some related observations about the limitations and opportunities presented to those attempting to devise and implement simple plans within complex systems.
Air travel as a complex system
Put in the simplest possible terms, a complex system can be described as a system that is characterized by a variety of components (often called ‘nodes’) as well as a high degree of interdependent connections between these components (often called ‘links’). Importantly, the different composition of components and their interdependency means that changes to one component can cause a ‘ripple effect’ throughout the system. These ripples are often referred to as second, third, etc. order effects. Emergence, self-organization, openness and adaptation of both the components and the system itself are also prominent characteristics. (For a thorough yet accessible introduction to complex adaptive systems, which provides much more detail than I can herein, readers are encouraged to consult Melanie Mitchell’s Complexity: A Guided Tour).
Picture 2: Example network diagram showing nodes and links. Source: US Army FM 3–24
The structure and processes enabling international commercial air travel present an excellent example of a complex system. At the most basic level, airports can be considered as nodes in the system and the flight routes between airports can be considered as links. This is illustrated in Picture 3, which shows only Qantas international freight shipping routes. Add to this passenger routes and domestic routes, and the picture of a complex system starts to emerge. And that is only for one airline — at present there are over 500 international airlines in operation. Factor in all of these airlines and the airports they use and it is clear the system is inherently complex — and very, very large.
Picture 3: Qantas international freight routes. Source: Qantas Freight
Yet there are also several other systemic characteristics that contribute to the complexity of this system. These include regulatory frameworks for matters such as the use of national airspace and air safety; the airplane manufacturing industry and the many companies that constitute it; flight engineering and resultant aircraft capabilities and limitations; and the economics of financially viable air travel, to name just a few. For the purposes of the example of my own travel experience given herein, however, it is sufficient to examine only the functioning of airports and flight routes, considered exclusively as nodes and links.
Nodes can be more or less ‘central’ within this system, with centrality being determined by one or more relative factors such as number of links or frequency of causal relationships to other nodes, rather than by their physical location. For example, large international airports with a higher rate of flight arrivals and departures to and from a relatively higher number of other airports may be considered more central; whereas small regional airports hosting only a few flight arrivals and departures to a few other airports may be less central.
Links could also potentially be considered as relatively stronger or weaker, depending on the frequency of flights along each route between two airports. This characteristic has led to the categorization of international air travel as a temporal system, a type of complex system wherein the links between nodes change over a given timeframe. This is shown in Picture 4. The temporal dynamics of this system are particularly important to this case study because, according to Luis Rocha, temporal dynamics matter ‘if one wishes to understand the cascades of delays caused by unexpected weather conditions during a single day’. This is because temporal mapping of the network can enable it to be studied as a series of paths between nodes that open and close at certain times.
Picture 4: Airports shown as a temporal network. Source: Rocha (2016)
From a passenger’s perspective, this is usually how the system is experienced. In practical terms, this system characteristic means that a change to conditions of one node (for example, delays to flights departing at one airport) can cause a ripple effect that alters the conditions of several other nodes, even those nodes that are not directly connected to the first (to continue the example, delays to departures cause delays to arrivals, which lead to further delayed departures across several other airports, and so on). This is certainly how I experienced the system during my recent journey to Canada.
Getting to Canada: A twenty-first century saga
It feels like only a slight exaggeration to say that, like Conan in his youth, mine is a tale of woe. The first unplanned event on my journey was small enough, and a fairly common occurrence that most readers have no doubt experienced. A storm at Sydney Airport temporarily closed its runway, causing a backlog of delayed departures and arrivals. My flight was one of those affected, the result being a delay to my Saturday morning Adelaide to Sydney flight by approximately 90 minutes. The flow on effects of this short delay would eventually become massively disproportionate, in keeping with the characteristics of complex systems, but when I arrived in Sydney I had not yet realized just how disproportionate they would be. Nor could I have realized at this point…..
This article was originally published on Medium. To continue reading the full article, please click the following link: https://medium.com/@aaronpjackson/your-flight-is-delayed-the-impact-of-complexity-upon-a-simple-plan-a2f70d31281