As a frequent traveller, one looks for and takes advantage of any program that will reduce the amount of time spent in airport lines. One such program is a trusted traveller program that pre-screens travellers, allowing for quicker security scans and expedited travel when crossing the border of say Canada and the United States. At my home airport, normally this would take 15 – 20 minutes at the busiest time of day on the busiest day of the week. It was different this week- it took almost an hour to navigate my way through security and Customs, which was fine given that I had allotted extra travel time to do something I so infrequently do: check my bags.
I don’t know what the root cause of the delays were this particular morning. Neither it seemed did the staff at the security checkpoints, or those trying to manage the expectations of the exponentially growing masses of confused and time-constrained passengers. I am willing to bet that only the population of tribbles could grow at a faster pace than the number of people joining these lines on this morning. As I slowly zig and zagged up and down formally defined lines, and several ad hoc ones as well, I wondered what might be happening, thinking about what could be done differently. What if the airport had a more predictive way to deal with a sudden influx of passengers arriving at the airport, or a model that shows the impact a slow-down at the Customs checkpoint has on the operational efficiencies at the security checkpoints? As the weather was clear, and the number of departing flights were constant with other Mondays, what other information would have been useful to move the hundreds of carry-on bags and their owners through the lines?
Surprisingly, there is considerable academia devoted to the science of lines, focusing on how to make them as operationally efficient as possible. One of the key principles in the design of lines (there are several principles) is to provide an unambiguous concept of where a line starts, where it is heading, and where it ends. This principle was clearly lost leading into and past the security checkpoints, as lines formed at will between the x-ray machine belts and the metal detectors. It did not help that some security officers were trying to move these lines out of these areas whereas others were encouraging people to form lines in these same spaces.
Since airports necessarily have to follow the single-line-to-multiple servers design model, the challenge thus to be addressed is to predicatively determine in relatively real-time when lines will become longer, and how these lines will proceed based on the resources at the end of the line (in this case, Customs agents). There are a number of data points already available at the airport – every passenger has to check-in, and when they do so at the check-in desk or kiosk, it is reasonable to assume that the next step they will take is to proceed to the security gates. The number of at-airport, check-in units moving from one point to the next is a relatively known number.
However, many people use mobile platforms to check-in, and do so upwards of 24 hours in advance. Their entrance into the airport is an unknown, is highly variable and thus subject to approximation. It is reasonable to assume that if 1500 people are booked on flights within the next 90 minutes, 1500 people will pass through security gates and say hello to the Customs officials. There is still a guessing game happening though. What if something delays many of these 1500 people from getting to the airport in a staggered fashion, and instead all arrive at once? It happens…
There are avenues that can be used to reduce this guessing game. For example, the tickets issued at the entrance of the parking garage can be an indicator as to how many people are about to enter the airport. The passenger volume on inter-airport trains or shuttles are clues as to where people are heading. I don’t know if this happens anywhere, but imagine if mobile airport passes could activate an arrival acknowledgement when the passenger is entering the airport – this could determine the volume of people about enter a workflow chain (the workflow being the security and/or Customs lines). With the predictive ability to determine what the volume of the workflow will look like, the proper resources can be scheduled and distributed accordingly. Using proper tools to collect this information, should there be any unanticipated changes to traveller behaviour that could alter the workflow, resources could be reallocated in real time to effectively deal with this change. Once through, these 1500 passengers will have the time to buy items before their flights.
Big Data, IoT and Digital Transformation are emerging and fascinating themes in business today. Irrespective of what the particular business product or service offering is, the goal is increase more revenue and profit over their competitors. Major airports are no different: it is a stated goal of my home airport to increase the number of international travellers passing through it, which will ultimately increase revenue. By increasing the number of travellers coming through its doors means these people are not laying over at another airport. More coffee and stereotype memorabilia is sold. If the airport is unable to show it can deal with increased volume of people arriving to begin with, then carriers may select another location to pass through.
Fortunately my bags made it to my final destination. I wonder what next week will be like when I repeat the process all over again. I will take the extra time out of my day to get to the airport early, check my bags, and proceed to the security lines and then Customs officers. What the airport will not know is when I plan to arrive, because I will check-in online and not notify in advance that I plan to check bags…