Timo Rissanen, head of Helsinki ground experience at Finnair, and Tuğberk Duman, project manager for Futurice, share some of their findings from a facial recognition check-in trial currently taking place at Helsinki Airport in Finland
Facial recognition technology is an increasingly familiar element of the airport experience with automated face recognition kiosks replacing passport control personnel at a growing number of airports globally. An obvious next step is to explore how face recognition could apply to the check-in process and other service touchpoints, and to individual airlines’ systems and procedures. Finnair, airport services provider Finavia, and digital innovation consultancy Futurice came together to design a prototype for the world’s first face recognition check-in desk system currently being trialed at Helsinki Airport. The system allows 1,000 Finnair frequent flyers to upload their photo via a dedicated test app before arriving at the airport. When checking in to their flights, the customers use a designated check-in desk where face recognition technology means check-in staff can recognize and greet them by name, without the need to produce any documents.
Although the trial is still underway, the thinking and insights behind the experiment revealed four key lessons for other airport service providers and airlines thinking of deploying face recognition technology:
1. Balance the need for operational efficiencies with the customer experience
The complexity of moving customers and luggage swiftly through the terminal, and the need to eliminate delays, explains why to date, airport face recognition innovation has largely fixed on improving operational efficiencies. This can involve inconvenience for customers who are often expected to add an extra element to their journey by stepping into a kiosk to register their arrival.
Our face recognition test aims to balance operational efficiencies – whether face recognition could help speed up the check-in process – with a desire to harness this innovative technology to benefit customers. Specifically, our aim is to focus on premium frequent flyers and to reward their loyalty by involving them in an interesting experiment, which if successful, will make check-in invisible by recognizing customers on the go. Frequent flyers are also an easy group to work with, as we already know a fair amount about them so they are a good test market.
2. Don’t replace staff with tech: use tech to enhance what staff can do
There is a growing tendency for face recognition solutions in airports to focus on replacing human-to-human interaction with human-to-machine interaction, in check-in, security and boarding. We wanted this experiment to be human-centric and to explore how tech can enhance employees’ roles rather than replacing them. One of the test’s objectives is to explore whether face recognition technology can make the job of Finnair check-in staff easier or not, and allow them more time to focus on the customer, rather than the documents. For example, if the system works correctly, check-in staff should be able to greet the approaching customer by name as their photo, identity and their journey details will have flagged up on the computer. As well as being more efficient than waiting for customers to produce their passports, this type of interaction makes for a warmer, more personal engagement for customers and staff.
3. Question the need to invest heavily in new hardware/software
Airport face recognition kiosks require a heavy, up-front investment in technology; however this experiment explores a solution focused on smartphones. One of the issues we wanted to address is whether it’s necessary for airport service providers and airlines to invest in expensive hardware when cell phones can do a similar job and nearly all customers have one. The system deployed for the test was built using widely available hardware and software, as well as cloud-based services, with a strong focus on the customer experience.
4. Have robust measures for protecting customers’ personal data
Asking customers to download an app, take photos of themselves and upload them to an airline’s server involves trust on both sides. Finnair demonstrates trust in its frequent flyers by delegating a part of the check-in process to them and trusting them to take their photos correctly. Meanwhile, customers trust the airline with their photos. Face recognition experiments need to have foolproof systems in place to safeguard security and to protect customers’ personal data.
To ensure our test is secure and to check the face recognition system is working, once customers taking part in the experiment have been identified using the face recognition system, Finnair staff physically check customers’ passports and boarding passes. The Finnair/Finavia/Futurice experiment has also been rigorous in protecting customers’ personal data by immediately converting the photo a customer takes into an untraceable biometric ID stored in their personal folder in Finnair’s database. The customer’s photo is then discarded. When the customer arrives at Helsinki airport and approaches the dedicated check-in, the face recognition camera takes their photo, confirms that there’s a face in that frame, and compares it with the biometric ID on our database. If the biometric IDs match, we have identified our customer.
In conclusion, face recognition is part of a larger biometric megatrend and is likely to become mainstream in the coming months and years. As well as making air travel flow more smoothly, airports and airlines need to ensure that the human customer and employee experience is at the heart of innovation and experimentation involving this groundbreaking technology.
May 18, 2017