Deep Turnaround: Predictable turnarounds and improved collaboration
Deep Turnaround improves the aircraft turnaround processes based on historic, real-time and predictive insights for all stakeholders. Using AI image-based processing, the Deep Turnaround algorithm detects and reports over 70 unique turnaround events in 30 turnaround processes. It detects delays as early as 40 minutes before the targeted off-block-time, and helps to make informed decisions.
Improve On-time performance
Deep Turnaround gives insight into what before was a blind-spot in data. Historical insights provide the opportunity to change processes structurally.
Make turnarounds predictable
Based on over 150.000 turnarounds , Deep Turnaround’s predictive capabilities detect delays before anyone else can.
Increase collaboration
Effective communication is key for effective airport processes. Objective turnaround data facilitates a base for constructive discussion and cooperation.
No great turnarounds are done alone – insights bring everyone together
Use the filters in the Deep Turnaround dashboard to show the flights that are critical. The dashboard shows the Predicted End of Ground Handling Time (PEGT), it shows the current status of the turnaround process and it shows when delays are expected considering the Target Off-Block Time (TOBT). You can also zoom into any flight to see the actual turnaround status, time stamps, handler and original planning.
Become a smooth, turnaround, operator
You can find all flight information in the dashboard: registration number, aircraft type, ground handler, airline, relevant time stamps, etc. You can also view the Predicted End of Ground Handling Time (PEGT), including a confidence interval.
Collaborate with airlines and ground handlers to optimise turnarounds
Turnaround insights can be used in two ways. They can be used during day-to-day operations, to help ground handlers allocate people and equipment and set accurate and realistic Targeted Off-Block times (TOBT). Gate planners benefit, knowing exactly when gates are available for incoming flights so they can reduce potential conflicts.
They can also be used to improve the operations. Historical data of all turnarounds can be combined to understand what needs to be done to improve the turnarounds, together.
The Turnaround process – a source of delays and misalignment
The turnaround process, at the heart of airport processes and yet a blind spot in terms of insight. Every day, this is causing delays, capacity inefficiencies and posing challenges in collaboration with sector partners.
At most airports, between 40-50% of delays occur during the turnaround process. Delays usually lead to more delays: delays in the turnaround of one aircraft can lead to a last-minute gate change for another, which again could cause delays in ground handling leading to a missed runway slot. Throughout this process, there is a lot of misalignment between all involved sector partners, with different perspectives of the truth.
The value of Deep Turnaround
Optimal slot usage
Runway slot allocation is used in congested airports to share scarce airport resources. Runway slots can be missed due to turnaround delays, waiting for the next available slot can take a long time. Deep Turnaround gives decisionmakers timely information on predicted delays. This allows airports and airlines to optimise available airspace, reduce the duration of delays, and improve on time performance by up to 2%.
More efficient asset & workforce utilisation
Resources are scarce in the turnaround process and idle time is costly. Yet it’s hard to plan realistically, as operations coordinators often have limited information on all current turnarounds. By providing real-time and predictive information about the turnaround process, Deep Turnaround allows you to plan more realistically and replan when necessary. Improving asset usage, resulting in lower idle time, and lower peak pressure for employees.
Fewer last-minute gate changes
Looking forward to fewer airplanes with long buffer times and fewer last-minute passenger flows in the terminal? The predictive features of Deep Turnaround can reduce the need for last-minute gate changes by 25-50%. It does so based on the availability of more accurate and timely information to initiate a gate change, as well as better-informed decision by gate planners to buffer an aircraft. This directly impacts on-time performance.
-
Case study
An in-depth look at the problem, the solution and the results.
-
Stay connected
Follow us on LinkedIn for more updates on Deep Turnaround.
-
Contact us
Get in touch with the Aviation Solutions team. Find more details, here.
Why choose Schiphol?
At Schiphol, Deep Turnaround is considered a foundational technology, imperative to achieving our strategic goals. Over the years, technological advancements and increased interest from sector partners, have driven Schiphol to continuously improve the Deep Turnaround solution to be faster, more efficient, accurate, while being easy to use.
Airport experience
As an airport, we know what it takes to make this happen. Insights alone don’t bring you value. You will benefit from our learnings and expertise to ensure value is created from the start.
Unique data model
Our unique data model and technical capabilities bring you accuracy, efficiency, maintainability and experience, with limited required calibration at your airport.
End-to-end delivery
You want a solution that delivers results. Schiphol and its partners, PA Consulting and Microsoft, bring years of tech implementation and run experience to the table. We’ve got you covered, end-to-end.
Choose the right product for your airport
Using AI image-based processing, the Deep Turnaround algorithm detects and reports over 70 unique turnaround events in 30 turnaround processes. After coupling this data to flight data, a second algorithm comes into action. Having learned from other turnarounds, it is able to detect delays as early as 40 minutes before the targeted off-block-time, allowing for informed decisions to improve processes.
Finally, the insights are conveniently made available to you through a dashboard and/or data stream. We offer three type of packages of our Deep Turnaround solution:
| Deep Turnaround Insights Dashboard |
Deep Turnaround Data Stream |
Deep Turnaround Data Stream & Turnaround Insights Dashboard | ||
|---|---|---|---|---|
Access to the turnaround insights dashboard, including:
|
Access to a real-time data stream of all turnaround events, including:
|
Access to the dashboard and the data stream, including:
|
Here’s how to make this work at your airport
Once you’ve made the decision to implement Deep Turnaround at your airport there are a few things to get done. Here’s how we make sure Deep Turnaround delivers the value you expect it to at your airport.
Step 1: Preparation
First, we determine which package (product, support module, add-ons) is most suitable for your airport. In addition, we will jointly develop a camera plan for your airport, allowing a clear view of the aprons in scope, and we validate the technical requirements.
Step 2: Deployment
During deployment, cameras will be mounted on site (with your local contractor), The full system architecture is deployed including any physical or virtual infrastructure and integrations with input data is established. The AI model will be calibrated and tested.
Step 3: Validation and go-live
The system is made live in a production state and all stakeholders are offered role-specific demonstrations and training. To make the deployment of Deep Turnaround a success, user onboarding is key and is always included.
Step 4: Support
During the first months we monitor performance closely and improve the predictive model based on the first data collected at your airport. In addition, we ensure users and stakeholders get confident in using the solution and improving the processes.
Frequently asked questions
Who are the typical users of the Deep Turnaround solution?
At Schiphol, the main users of the Deep Turnaround solution are Airport Gate Planners, Air Traffic Control, Ground Handlers, Flight Officers, Apron Controllers and Operational Excellence Managers. In the future, we expect to include features which can benefit safety and sustainability officers/managers as well.
How many turnaround events are included in your model?
There are 70 turnaround events included in our model, giving insight in over 30 turnaround processes. If you choose to implement Deep Turnaround at your airport, all turnaround events which we have access to at Schiphol are included automatically.
Is Deep Turnaround suitable for my airport?
Deep Turnaround is suitable for most airports, small and large, and irrespective of apron design. Whether you plan to integrate Deep Turnaround in your own dashboarding using the data stream, or whether you plan to use the Turnaround Insights Dashboard. Deep Turnaround is especially suitable for capacity constrained airports.
Can Deep Turnaround work with our own cameras?
Yes, under some conditions it is possible to use existing cameras. Cameras have to be IP cameras, as the Deep Turnaround solution needs to be able to connect to them through an HTTP(S) protocol. Furthermore, there are some requirements for how the cameras should be positioned on the apron (approx. 20 meters high), as well as for image resolution with a minimum of 1280 x 720 pixels.
How does your product compare to other products on the market?
We work with one end-to-end model of all turnaround events, instead of different models for each turnaround. This makes the model easier to maintain and more cost efficient. Also, our model works on the principle of sharing data, if you select the Deep Turnaround solution you opt-in that your data will be used to train the model. Similarly, you can benefit from the model that was trained at Schiphol Airport. In addition, our solution is created at Amsterdam Airport Schiphol in collaboration with users, this allows us to make all Schiphol functionalities, now and in the future, available for other airports as well.
How long does it take to implement the Deep Turnaround solution?
As our Deep Turnaround solution uses a single model approach, wherein one AI model learns more every time it processes an image, we do not need to start from scratch at your airport. The algorithm instead can build on what it already knows. From the moment the cameras are installed on the ramps, implementation is expected to take around six weeks. That includes setting up and configuring the Deep Turnaround system, integrating different data sources into the Deep Turnaround algorithm, performing security assessments, and training users.
Partners
