Flight delays are a common occurrence for many. According to the BBC, 11% of flights were delayed in the second quarter of last year while 16% of US flights took off late in 2021 as a whole. Flight delays can also be irritating for many. If a flight is delayed, it means one has to wait even longer in the waiting area. For most international flights people are expected to allow 3 hours for checking in. In most cases checking in happens quickly and a lot of time is spent waiting for the flight. For a lot of holidaymakers, this may be seen as part of the holiday as many may have a meal in one of the many restaurants in the airport or enjoy duty-free shopping.
However, for those travelling on business, apart from the pleasure of using the amenities or having gourmet cuisine in the business class lounge, a delay could be potentially more disruptive for their work. Flight operators may also bear the brunt of paying compensation to customers and extra fees to the airport if they are spending more time on the ground. So delays not only cost time, they can also cost money to both the passenger as well as the flight operator. So how can we prevent this?
In order to come up with a solution, we would first need to understand the causes of delays in the first place. According to the BBC, there are often multiple reasons beyond delays. These could include ground crews running late with aircraft refuelling, bad weather, having to wait for their pilots, or the airport struggling with the sheer volume of passengers. Some of these factors such as bad weather may be harder for a flight operator to account for. However, one strategy to reduce delays in other areas is the increased use of high tech monitoring technology, to help airports and airlines to run more efficiently.
One such provider of hi-tech monitoring is the Israeli start-up IntellAct. According to the company, most delays result from problems with what they call turnaround services, which are within an airline’s control. These include an airline being refuelled and restocked (with food for the passengers for example), or luggage being removed or reloaded.
IntellAct’s artificial intelligence (AI) software system can automatically detect delays in these turnaround systems. This AI would be easier to introduce as it would work using existing security and observation cameras of airport operators and airlines. After detecting any delays, it can then highlight the problem to airport staff and ground crew and suggest a mitigation plan.
For example, if a truck arrives late, IntellAct can detect this and recommend a system known as parallel boarding. This is where passengers board while an aircraft is being fuelled. This however does require a fire engine to be present as a safety precaution.
IntellAct says that its technology is currently being evaluated by a number of airports in Europe, Asia and the Middle East. Preliminary tests conducted by Israel’s national carrier EL AL at the country’s Ben Gurion Airport near the town of Lod (9 miles southeast of Tel Aviv) have shown that implementation of the artificial intelligence from IntellAct could potentially curt turnaround times by 15%.
There are however some drawbacks. Even with the AI installed, as shown by the example of parallel boarding described earlier, it appears that because of the delay a fire engine would now need to be present during boarding in order to mitigate it. However, whether it would be worth spending the extra money on having a fire engine present would be something that would need to be weighed up and decided by the airline in question. Therefore, whether the AI gives appropriate solutions that would justify implementing it is something that airlines would have to consider even if delays are reduced.
In addition, as suggested by Professor Sandra Wachter, a senior Research Fellow in Artificial Intelligence at Oxford University, having technology like this used may cause more stress to employees due to constant monitoring of the activities. As one can see more research may be needed before this type of artificial intelligence gets implemented. However, if implemented properly delays can be reduced.