Friday, March 9, 2018

Artificial intelligence in Airport?


Anyone and everyone who has travelled by Air knows the importance of time.   Airlines advice passengers to reach airport at least 3 hours before flight departure and for international flights and about 2 hours for domestic.  While passengers enjoy being in the retail area of airport but unfortunately most of this time is spent standing in queues for checking and security clearance. 

Globally, the average time spent at an airport — from arrival until an aircraft’s doors are closed — was 133 minutes in 2014. Out of this time only around 30 minutes is spent by the passenger having relaxing time in retail area. The time consuming process in airports include- Check-in/bag drop, Security, Orientation, and Boarding. This is a loose- loose situation for everyone. Passenger’s experience is bad as he has to wait in the queue. The airlines cost is more because they have to put extra staff to handle customers, the Airport security time staff also need to spend more time.

The best possible outcome for an Airport is when passenger spend more time in retail area, which would result in more sales in these area which in-turn in more revenue for airport.  Airports around the world have transformed into a high tech, heavily secured shopping malls to capitalize this opportunity.  Which the concept of duty free purchase, it give added incentive to passengers to purchase.

Many processes have been introduced that have reduced the queue time for passengers. Airlines provide option to passengers to web check-in and use home printed bag tags. When they reach airports they just need to drop their bag and proceed towards security.  In US the TSA, who is responsible for airline and airport security has introduced a pre-screening program called TSA Precheck. Through this frequent flyers can enroll to the TSA precheck program, on enrolling they would be issued a Known Traveler Number (KTN) after successful background check. When they reach airport they have a special queue for security clearance which is many times faster than regular queue. All these initiatives have helped reduce the time spent by passenger in queues, but with limited success. The problem is the TSA Precheck is not available to everyone e.g. non US resident, and not everyone wanted to go through the process of getting the KTN number.

Now coming to the main problem, how can airports process be streamlined to ensure all passengers spend more time in retail area then in queues? This is where technology and in particular AI and biometrics can play a major role. Many recent developments in AI technology has introduced algorithms that can detect human emotions, this enables the system to see if a person is stressed or scared. This technology is maturing fast and many leading players including Microsoft has adopted it. This would enable profiling of people with bad intentions. This could be clubbed with facial recognition and government database check to ensure the profiling is done properly. This would mean that only suspicious passengers go through extra checks and frisking and all others can basically clear security check in seconds.  This can go a long way in making passengers less hassled in airports. This saves time and cost for airlines also as they don’t have to send passenger information to government agencies for domestic flights as required by many governments. When passenger spends more time in retail area rather than in queues, it’s a win-win for everyone.

Although this technology is in early stages and would need manual interventions in the initial phase, with machine learning system would get smarted day by day and at some point would be able to handled security checks on its own in a fully automated way. The future is promising.

5 comments:

  1. It is worth giving it a try by having this simulated in a lab and look for champions to promote this idea.

    ReplyDelete
  2. As a air traveler, I can relate to the challenge. If AI is going to cut down the time saved in security process, I would certainly be happy. What could be the key hurdles in this journey? It will help to get some clarity. Once this idea becomes a reality, will I spend more time shopping? I don't know.

    ReplyDelete
  3. The use case that you covered clearly aligns with Unisys' niche. It'd be great to see what happens. The future is indeed promising!

    ReplyDelete
  4. Bright spot...There are examples like KTN to learn deeper and globalize it around the world. Also, facial recognition becoming commodity in smart phones. This can be utilized to reduce the load on AI further with some pre-checks.

    Tough spot...Who will invest in AI? Airports or Airlines. Who are the beneficiaries and to what extent is a difficult question to answer.

    ReplyDelete
  5. Well articulated. I believe this has the potential to be something bigger. Imagine if this is implemented, could this be a win for retail? I doubt so. I am sure I would want to turn up at the airport much nearer to the departure time if the checking process only takes me 10 - 15 mins.

    ReplyDelete