Eindhoven Airport recently implemented Deep Turnaround, an AI-driven tool designed to optimise aircraft turnaround processes. We spoke to Lennard Albarda, product owner, passenger & aircraft systems and Frédérique Portheine, manager operational excellence. They shared their experiences and the strategic approach they took to roll out this innovative solution, focusing on refining operations and building stakeholder trust.
The aircraft turnaround process is a critical yet often fragmented aspect of airport operations, executed mainly by airlines and ground handlers. Since Eindhoven Airport is mainly serving low-cost carriers (Ryanair, Wizz Air, Transavia), supported by two handlers (Viggo, Skytanking) short turnaround times are crucial. However, a lack of real-time visibility and predictive data made it difficult to guarantee short and predictable turnarounds at all times. Also, because of the airport’s size, inefficiencies can have a disproportionate impact, making it essential to find a solution that could be customised to its unique needs.
Deep Turnaround (see box) addresses these challenges by using AI-driven image processing to monitor and optimise apron activities, providing real-time and predictive insights that enable proactive management and improved coordination across all stakeholders. The tool’s success hinges on building trust with partners and gradually introducing transparency, ensuring that everyone is aligned and engaged.
6.8 Million passengers (2023)
Over 41,000 flight movements (2023)
Mainly serving low-cost carriers (LCC): Ryanair, Wizz Air, Transavia
Short turnarounds due to the LCC models
Two main ground handlers: Viggo, Skytanking
Eindhoven Airport first implemented Deep Turnaround at four stands in November 2023 before rolling it out to all the airport’s aircrafts stands in the summer of 2024. Developed by Aviation Solutions, the tool provides real-time insight into various turnaround processes, from passenger disembarkation to refuelling. Instead of a full, immediate rollout, Eindhoven deliberately implemented the system in stages to build trust and acceptance among stakeholders, while refining its use and ensuring smooth integration.
‘Normally, I would opt for an implementation all at once,’ says Portheine. ‘But by introducing Deep Turnaround in stages, we’ve found that it helps us focus on what we and our partners really need. This step-by-step approach improves collaboration and ensures that everyone feels involved in shaping the product, which builds trust along the way.’
It’s also important that Deep Turnaround is seen as a tool for collaboration rather than control. That’s why handlers were gradually introduced to the system and its capabilities, and their feedback was actively sought and integrated. This approach not only improved the accuracy and usability of the tool but also created a sense of ownership among the handlers.
‘The more people see Deep Turnaround in action, the more their imaginations start to capture what it can do, even if it’s not fully realised yet,’ Albarda says. ‘When it was just an idea, the benefits weren’t as clear. But once we presented the live dashboard, it came alive for them. Now, the possibilities are clearer, all operational staff at Eindhoven Airport is using it, and enthusiasm is growing.’
Before implementing Deep Turnaround, Eindhoven Airport faced significant challenges due to a lack of visibility into the turnaround process. ‘Much of the turnaround process was a black box for us,’ explains Albarda. ‘We only found out about problems after they had occurred. And these problems often have a knock-on effect on other processes.’ This lack of real-time visibility often led to delayed responses and inefficiencies in managing on-time performance (OTP).
Driven by the need for better data and proactive OTP management, the first phase of the Deep Turnaround implementation involved installing two cameras at four selected aircraft stands to capture real-time footage of the turnaround process. The AI system analysed this footage, providing valuable insights and provides real-time warnings of delays. ‘Capturing live apron footage over a longer period of time creates a database of historical data.It provides us with accurate information. And when we share that information with the handlers, we have a common truth which acts as the perfect starting point for improved collaboration,’ says Portheine, highlighting the importance of transparency and accuracy in the process.
The first results were promising. Eindhoven found that the AI system was highly accurate, with a reliability of between 96% and 98%. Encouraged by these results, the airport decided to roll out Deep Turnaround across all 14 of its stands, while continuing to explore how best to integrate the data into its operations. ‘The benefit of common truth is that you don’t have to look so much at who caused the problem, but what caused the problem. This adds to the feeling of having a shared responsibility and creates more willingness amongst all parties to improve the process,’ adds Portheine, emphasising that this shared understanding allows structural adjustments to processes, ultimately leading to better OTP management. For example, preliminary data shows that Deep Turnaround can help prevent over 80% of expired Calculated Take Off Times. Detailed results will be closely monitored.
The shared approach of Eindhoven Airport doesn’t only focus on the local stakeholders. ‘EUROCONTROL has the same problem we had, but on a bigger scale, they have no data about what’s happening on an aircraft stand. When a plane is in the air, they have detailed data about location, speed and direction. But once the plane lands, it all becomes a big question mark.’ We envision that predictions on the end of ground handling time can help EUROCONTROL optimise their network. Albarda explains. ‘By using Deep Turnaround’s Predicted End of Ground Handling Time (PEGT), EUROCONTROL can provide a more suitable slot time, because they have a much better idea about whether the planned departure time is feasible.
Portheine adds: ‘If we share this type of data with each other, we can make much better use of the airspace, in Europe and North Africa in our case. It increases the efficiency and can reduce the slot wait times. It’s a big promise but when we work together, we can achieve it. I believe the future will be one with cooperation, shared knowledge, and being transparent. It’s not about who’s at fault, but about collaborating as one, to optimise operations and achieve everyone’s goals.
While the initial focus of Deep Turnaround is on optimising turnaround processes, Eindhoven Airport is already exploring other applications for AI in their operations. For instance, they are considering integrating sound cameras to monitor non-visual events at airplane stands, which could provide even more comprehensive data on turnaround activities.
As Eindhoven Airport continues to refine and expand its use of Deep Turnaround, it’s clear that the system has the potential to significantly improve turnaround punctuality and On Time Performance. The airport’s experience underlines the importance of a collaborative approach, where technology is not just implemented top-down, but developed and refined in partnership with all stakeholders.
‘I appreciated the flexibility to introduce new ideas and adapt the tool to our specific needs,’ Portheine says. ‘Although we are a smaller airport with different requirements, we were given the space to make the tool fit our operations. This kind of tailored approach was a very pleasant experience for us.’
For other airports considering similar technologies, Eindhoven’s experience offers a roadmap: start small, engage stakeholders early and often, and be open to iterative improvements. Albarda highlights one of the key benefits of their chosen solution: ‘The fact that Deep Turnaround’s AI model makes use of the experience at multiple airports was a huge advantage for us. We could immediately use the huge amount of existing turnaround data, which gave us a significant head start. While others were building AI models from scratch, we saw strong results from day one.’ See also, “How does Deep Turnaround work?".
As more airports adopt AI tools like Deep Turnaround, the industry as a whole can move towards more efficient, transparent and ultimately more successful operations.
Deep Turnaround is an AI solution designed to optimise aircraft turnaround times by monitoring and analysing over 70 key events across over 30 processes, providing real-time insights and alerts. Used by airport gate planners, air traffic control, ground handlers, and other operational roles, it’s suitable for airports of all sizes and can integrate with existing IP cameras under certain conditions. The system uses an AI-based algorithm to analyse images, recording when activities start and stop, and predicting potential delays. This information is then displayed on the Turnaround Insights Dashboard, making it easy to identify flights requiring extra attention. The software implementation typically takes around six weeks, using a single, continuously learning AI model for efficiency and cost-effectiveness.
Discover what we're up to and receive a quarterly digest of everything related to Aviation Solutions.