How Schiphol Airport is tackling queues with data
At IoT Tech Expo in Amsterdam last week, Schiphol Airport’s chief data officer began his talk on data and aviation by letting international guests know that he can’t give individual guidance on how early to get to the airport, “but the general advice is four hours.”
For a one-hour flight back to the UK, that’s at least half a day’s travel.
With queues backing up for miles and passengers missing their flights as a result, Tor Bøe-Lillegraven, who has been chief data officer of Royal Schiphol Group since 2019, admitted: “It’s quite brutal.”
The passenger issues and airline delays at the Netherlands airport have remained unresolved for months, culminating in the resignation of Schiphol Airport’s CEO Dick Benschop, earlier this month.
The queues and missed flights caused chaos for many delegates from events such as broadcast tech fest IBC, with many left stranded and sleeping in the airport overnight, until the y could catch a departing flight/
While many frustrated passengers may have praised the move, it left Bøe-Lillegraven – who has worked on the airport’s data strategy for the last three years, feeling disappointed.
There are many moving parts to solving these passenger issues, Bøe-Lillegraven explained, and it will be a challenge for the next CEO, but he said that he intended to do his part with data.
Taking off with data
Schiphol airport had already been undergoing its own digitisation programme for several years, connecting its data with Internet of Things (IoT) devices and edge computing.
Bøe-Lillegraven joined Schiphol three years ago after working at online marketplace eBay, as “their data guy.” In comparison to eBay, the challenges at Schiphol are vastly different, Bøe-Lillegraven said, because at eBay the consumer is on their personal device, and “when you find something, you buy something.”
So, although eBay has a huge amount of data, “it’s fairly homogenous.”
At Schiphol however, he added that there was “there’s all kinds of stuff” to consider: Passengers, staff, trucks, IoT devices (scanners and sensors), buildings and airplanes – all of which carry data.
Fortunately, Schiphol was already “one of the most digitised airports in the world,” by the time Bøe-Lillegraven joined, and it had already started to analyse and use its data to make decisions – a decent inkling, Bøe-Lillegraven added, that the airport had a good level of “data maturity.”
However, he noted: “One of the challenges for me, when I joined Schiphol in 2019, was that I did not know what was going on under the hood.”
While Bøe-Lillegraven claimed that he could see how the digital world connected to the physical world in a way that he’d never seen before at Schiphol, he thought that the airport needed to modernise its data strategy.
He noted that since he joined in 2019, many terms and processes had become outdated. For example, Schiphol airport used to have a centralised datalake but has now moved towards a more decentralised ‘data mesh’ system, to enable data to be accessed more seamlessly by individual project teams.
According to Bøe-Lillegraven there were three priorities needed to transform the way it handled and analysed its data to make the passenger experience seamless.
The first he stated, was the need to automate – a challenge that teams are still addressing in 2022, particularly with the need for automated baggage and security processes.
“We don’t have enough resources, so I would say we still need to automate,” he says. ”We haven’t got as far as we had hoped – but that’s life,” he admitted.
Secondly, Schiphol airport needed to recognise itself as a platform: “We had a lot of discussions at Schiphol, asking, are we a platform? Well, we are sort of, metaphorically. Travellers and airlines all coming together in this one physical place…” he reasoned.
So, the airport also needed to build a platform strategy and rethink how the data is shared.
Now, Schiphol holds its data on a data platform where data can be shared, collaborated on and users can get data and give data as opposed to every data point being separated.
Thirdly, he said that Schiphol needed to use its data platform to identify different types of business decisions. “For instance, in resolving security issues, to ensure that passengers reach their planes in good time.”
This included using data on the edge and using data with artificial intelligence (AI) he added.
The departure lounges
Bøe-Lillegraven highlighted dozens of processing points that hold data within the airport. “It is hugely complicated and complex and has proven to be a key challenge for Schiphol,” he added.
When Bøe-Lillegraven first started working on the airport’s new data strategy, the team realised that the web of process points was “simply too much”.
“So, we went back and had to tell a different story,” he explained.
“Number one, it will be cloud-based, and it will be public,” he explains. “Number two, we will have something that we call data ingestion, which is how you get data into the actual data platform.”
So, the team went on to set the different chronological steps in a passenger’s journey from queueing, check-in, security, to boarding and they retrieved data from each of these points.
By combining all these data points, the data teams produce “business intelligence, which details how many passengers are going into security and how many people have picked up their baggage,” said Bøe-Lillegraven.
“From that, you can get accurate data in real-time,” and thus, in the future, the data will be able to provide passengers with the ability to plan their arrival time prior to their flight better, by being able to see how busy the airport is and how long queue times are via the Schiphol airport app,” he added.
The airport is also currently working with Uber to allow passengers to find out how early they should arrive to their airport if they choose to put their flight number into the app.
During Covid-19, it also had a heat map app within the airport, which tracked mobile devices that were on the app and allowed people to avoid busy areas.
In plane sight
During the quiet period of Covid-19 lockdowns, the data strategy team focussed primarily on tracking plane data.
When Bøe-Lillegraven first started at Schiphol, he learned that the airport didn’t have any real-time data on the activity happening with the movements of cargo, catering or baggage trucks around the planes.
“So, we installed 3D camera detectors and started doing live video feeds on everything that goes on around the plane.”
After the team manually annotated what each truck looked like, the information was then fed into an algorithm which could then recognise each truck on camera and keep track of it.
This took three years and has since helped people working in operations quickly realise why a plane might be delayed so that they are able to intervene more efficiently.
Subscribe to our Editor's weekly newsletter