Airside operations Turnaround’s black box
The turnaround process – covering the time from when an aircraft is parked until it departs, including ground handling activities like boarding and disembarking passengers, catering, loading and unloading luggage – has long been a blind spot in terms of insight, referred to as a ‘black box’ for stakeholders by Schiphol. Given the process’s importance within airport operations, this deficiency has historically led to delays and capacity inefficiencies – indeed, at most airports, 40–50% of delays occur during the turnaround process. And, as we all know, delays tend to have a knock-on effect, disrupting flights further down the chain. Airports typically see strong peaks in arrivals and departures at certain points of the day, leading to issues finding a gate for an aircraft. Gate planners and air traffic control (ATC) are responsible for addressing this problem, and typically possess limited insight into the status of ground-handling for these aircraft, which can lead to sub-optimal planning. These difficulties are compounded by the fact there isn’t a single source of data to provide turnaround information to stakeholders. Delays are often reported minutes before they occur, and even when an overview is available, the information is not always in real-time or up to date.
Schiphol began developing an autonomous solution as part of the group’s larger ambitions to have fully autonomous and data-driven airside operations by 2050. Deep Turnaround’s algorithm can detect and report over 70 unique turnaround events in 30 turnaround processes, and can identify delays up to 40 minutes before the targeted off-block-time. Deep Turnaround is currently fully operational across 65 connected stands at Amsterdam Airport Schiphol, with plans to scale-out across all the airport’s 127 stands in 2024. Cameras have been placed on the left and right sides of the stands, and take images of the aircraft every few seconds, which are then fed through Deep Turnaround’s algorithm. This sequence of images will tell the algorithm when a process starts, how long it goes on for, and when it ends – and this information is then displayed on the Turnaround Insights Dashboard, providing real-time progress of turnaround operations on each stand. The system also tracks timestamps that mark the end of ground handling time, which can help gate planners determine whether a flight that is still occupying gates will depart soon or is ready to depart. Similarly, for ATC, Deep Turnaround can inform them if runways slots are available for upcoming flights, or whether flights are due to depart but aren’t yet ready to go. For ground handlers, the algorithm can optimise resource allocation when it comes to pushback trucks or de-icing equipment. In the past, gate planners, ATC and other stakeholders would have had to use binoculars from the control tower, or CCTV cameras not intended for that purpose, to determine the real-time status of turnaround operations. “We streamline and
Future Airport /
www.futureairport.com
make the process of decision-making in real-time faster,” notes Jeffrey Schäfer, process owner, aircraft turnaround at Royal Schiphol Group. Deep Turnaround also provides predictive analysis for turnaround times, which can be hugely useful when it comes to resource allocation over runway availability, take-off slots and gate planning, as stakeholders can be alerted to potential outbound delays earlier than ever. In the past, much of this key decision-making relied on the expert judgement and experience of those working in airside operations, rather than hard data. “They know if fuelling is still taking place that it will take about 20 minutes, and by their estimates that will be the last process to finish,” says Schäfer. “But the algorithm sees beyond this – all the different things. For example, it might see that the catering has not been done yet. But it’s a long-haul flight, and we’re pretty sure catering needs to happen for this flight. So, that will also be incorporated.” Deep Turnaround also provides post-operations analysis to find ways airports can improve processes in collaboration with other stakeholders. “Delay codes, of course, have always been the main source for airlines and handlers to operate on, but they are, quite frankly, often based on opinions or best estimates, not giving a full overview of what happened and what was the root cause of a delayed flight,” Schäfer adds. “Using data we gather from Deep Turnaround, we can really pinpoint where the delay in the handling process can come from, and also look beyond what happens directly.” Finally, it also overcomes the challenge that has plagued airports over the transition of information between key parties. In the past, personnel were needed to manually track and enter the turnaround information, relying on their experience and judgement to make decisions. “Of course, that can work,” Schäfer notes. Indeed, it’s how the industry has operated for the past century. “We’re not doing this because we don’t have trust in our staff, but to release them from
Above: The Turnaround Insights Dashboard provides stakeholders with real-time data about what is really happening on stands and runways.
Opposite page: AI tools such as Deep Turnaround can enhance efficiencies across many airside operations.
40- 50%
82%
The accuracy provided by Deep Turnaround straight out of the box.
Royal Schiphol Group 27
The percentage of delays in airports that occur during the turnaround process.
Royal Schiphol Group
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37
