One of the most studied transport hubs is certainly the function of the airport. Within short time slots, large numbers of passengers are transferred from one plane to another plane or other modes of transport. Furthermore, the airport is only for few users an origin or destination, thus a hub.
When planning new airport infrastructure or adaptations in processes and technologies, companies and governmental organizations run various simulations predicting passenger flow under low and high utilization scenarios, emergency simulations, utilities planning etc… In general, this is done very well to avoid costly immediate effects (missed, delayed flights) and long term effects (capacity constraints, high maintenance costs).
However, once in a while I come across fascinating examples that show that sometimes people avoid planning and create with this a huge mess…
Overall airport capacity constraints
I experienced a fascinating example when visiting Abidjan last year, where the daily Air France A380 flight with its 516 passengers blocks the Port Bouet Airport for hours. As it is a well known fact that boarder and security control cannot handle the workload, people arrive very early and queue more or less patiently.
Given the building’s spacial constraints, there are not many options than speed up the process to remove the bottlenecks during the security checks or divide the checks in multiple process steps (such as early ticket check).
Last weekend however, I experienced something new to me. The obviously artificial bottleneck at the border control of O.R. Tambo International Airport in Johannesburg.
With the introduction of bio metrical scanning of passengers, South African immigration is definitely improving the quality of its boarder controls. While it is fully understandable, that rigor checking is established when entering the country, South Africa has obviously also put strong emphasis on transferring and departing passengers.
In the transfer area fo the airport, they have even displayed the actual time it takes now a boarder agent on average to inspect a passenger 2 minutes. As I had ample time while waiting in the queue, my measurement with a sample of 10 pax resulted in an average of 2minutes 44 seconds.
In the transfer area, they had 4 booths open, resulting in an hourly capacity of roughly 120 passengers per hours. With an annual number of passengers of roughly 20 million, of which about 9 million are international passengers, this capacity is insufficient and resulted in large delays, long queues in areas difficult to manage in a safe and efficient manor.
The obvious solution would be to improve the bio metric checking time or to add additional boarder agents – for both seems to be ample space, respectively process optimization would allow it.
So how to avoid bottlenecks?
The answer is very simple – run the obvious use-cases, simulate them and analyze possible bottlenecks and constraints. After doing so, develop scenarios to avoid this.
I just learned that there is actually a very simple solution to analyze the passenger flows and adapt the capacity of the different processes dynamically to the passenger flow. Xovis has developed a stereoscopic sensor and the software to identify passengers and analyze the flow of persons in a queue. This way, the operations can immediately react and prepare when they see more passengers coming into the terminal at the start of the process or also if passenger flow is lessening. Seems to be simple and very efficient…