ASTAIR, a SESAR JU project, is working to achieve the full automation of airport ground operations, enhancing both efficiency and predictability through an AI-driven support tool. 

The solution which the consortium will develop incorporates automated taxiing, seamlessly integrating with A-CDM (Airport collaborative decision making) and A-SMGCS (Advanced surface movement guidance and control system). It aims to reduce operators’ workload and delays, and optimise the operational flow while maintaining a human-centred approach. By involving ground operators in collaboration with AI at varied levels of automation, ASTAIR is poised to revolutionise how airports handle vehicle movements on their surfaces.

A key component of ASTAIR’s iterative development life cycle is validation with relevant stakeholders, such as domain experts and expected concept users. ASTAIR will validate the concept through different techniques; workshops, Real-Time Simulation and Fast-Time Simulation, within the scope of three validation exercises. These will evaluate different aspects of the concept and refine the solution based on feedback from airport operators and industry experts.

Validation Exercise #01, the focus of this article, is vital to the project’s initial assessment phase. Through three workshops held across major European airports, the consortium sought to define the scope and use cases for ASTAIR, validating its concept and expected impact on ground operations.

Validation Exercise #01: workshops and insights

The first ASTAIR's validation exercise comprised a series of workshops held between the end of 2023 and mid-2024. These sessions were aimed at gathering key insights from stakeholders, including airport operations managers, ground handling specialists, and autonomous systems experts from three of Europe’s largest airports: Paris Charles de Gaulle (CDG), Amsterdam Schiphol (AMS), and Frankfurt (Fraport).

The workshops provided an opportunity to discuss the current operational landscape, identify areas for automation, and assess the potential impact of ASTAIR on day-to-day airport operations.

1. Paris CDG Workshop – December 2023

The first workshop, held at Paris Charles de Gaulle Airport on the 18th of December 2023, set the foundation for ASTAIR’s validation journey. This session involved interviews and discussions with local stakeholders, focusing on their current procedures and identifying areas where automation could streamline ground operations.

One of the key insights from Paris CDG centred around the role of automation in specific operational scenarios. Stakeholders discussed the potential for ASTAIR to reduce operator workload by handling routine tasks like taxiing and vehicle coordination autonomously. The discussions also highlighted the importance of maintaining flexibility, allowing operators to tweak the system’s algorithm to accommodate unexpected operational changes.

2. Fraport Workshop – April 2024<

On the 19th of April 2024, the second workshop took place at Frankfurt Airport (Fraport). Like the Paris workshop, this session involved detailed discussions with airport stakeholders about the potential applications of ASTAIR’s technology.

The Fraport workshop focused heavily on use cases for automation, such as handling arriving traffic without parking and modifying runway operations in response to real-time conditions. One challenge identified was human-automation teaming, which emphasised the importance of balancing autonomous processes with human expertise. Liability concerns, particularly around path planning and decision-making, were also discussed, as were the implications of automation failure and system adaptability in high-pressure scenarios.

3. Expert Group Workshop – May 2024

The final workshop, held on the 24th of May 2024, brought together ASTAIR’s expert group for an in-depth analysis of the use cases identified in the previous sessions. This expert judgement analysis was pivotal in refining the use cases and addressing the complex challenges associated with path planning and human-AI collaboration.

The refined use cases explored scenarios ranging from routine taxiing operations to more complex situations, such as automation failure and runway configuration changes. The group also examined different levels of automation, categorising each use case according to the European Aviation Safety Agency’s (EASA) established levels of automation.

Key use and validation exercise

The workshops resulted in the identification of eight primary use cases (UCs), which will serve as focal points for ASTAIR’s ongoing research and development. These use cases provide concrete examples of how ASTAIR could improve ground operations through automation. The use cases include:

1. Normal operation: normal operations of inbound and outbound traffic taxiing on a low to normal load. Derived in two sub use cases, with and without the use of tow tugs:
   
   1. Departure with taxibot: automating the process of assisting departing flights with remote-controlled taxibot.
   2. Arrival with taxibot: normal arrival of an aircraft using a taxibot.
2. Normal operations with re-scheduling: managing routine traffic while accounting for real-time rescheduling needs.
3. Arriving traffic without parking: handling arriving flights when parking spots are unavailable.
4. High-level taxi strategy tuning: fine-tuning the AI routing strategy based on operational conditions.
5. Automation failure: ensuring human operators can swiftly intervene when AI is no longer able to compute a conflict free solution.
6. Runway mode of operation modification: dynamically changing runway mode based on current conditions.
7. Departure remote holding: facilitating remote holding solutions when the aircraft is ready to depart but departures are delayed.
8. Arriving flight with technical issue: managing ground movements when an arriving aircraft experiences a technical fault at landing.

Following the insights gathered from Validation Exercise #01, the next phase focused on enabling end users to experience how ASTAIR’s system may work through prototypes and demos.

Building on the insights gathered from Validation Exercise #01, the next phase focused on allowing end users to experience how ASTAIR’s system could function through prototypes and demonstrations.

Validation Exercise #02 comprised two structured workshops designed to gather feedback from end users, specifically Air Traffic Controllers (ATCOs). These workshops were held in June and July 2024, concentrating on validating the solution design and the quality of the information provided. The primary objective was to enable the ATCOs to interact with a simulator of the ASTAIR system. Not all the UCs were validated during this exercise, due to time constraints, but only three of them: departure and arrival with taxibot, normal operations with rescheduling and arriving flight without parking.

The outcomes of these workshops initiated new cycles of concept development, refining the use cases based on stakeholder feedback. Following a design walkthrough methodology, ASTAIR’s partners facilitated discussions after each interaction to collect reactions and pose specific questions related to situational awareness, such as the visibility of information, teamwork (both with and without AI), and task distribution within the system, among other elements.

The two workshops conducted during this second validation exercise yielded feedback that will be used to further advance concept development. Consequently, it can be stated that after each workshop, a new cycle of concept development was initiated. New ideas and requirements emerged, and the use cases were shaped significantly by the insights gathered from stakeholders.

The qualitative feedback from this exercise will be compiled and reported in the D5.2 Exploratory Research Report (ERR).

The use cases will guide future development efforts, influencing key work packages, including WP2 (path planning algorithms) and WP3 (human-automation interaction tools). While some challenges, particularly around scalability and system adaptability, remain, the workshops provided valuable insights into how ASTAIR can be tailored to meet the specific needs of different airports.

Impact on future activities

The feedback from Validation Exercise #01 and #2 will directly inform two critical project’s work packages. WP2 focuses on designing AI algorithms that can autonomously manage path planning and fleet coordination on the airport surface. The workshops highlighted several challenges, such as handling sudden operational changes and ensuring the system remains effective in high-traffic scenarios. As real-time adaptability is crucial for the success of ASTAIR, WP2 will need to address these challenges, ensuring the algorithms can provide feasible solutions even in complex, time-sensitive situations.

WP3, meanwhile, is concerned with human-AI interaction. The workshops underscored the importance of creating a user interface that allows ground controllers to collaborate effectively with the AI system. One particularly challenging area is Level 2B automation (EASA classification), which involves complex human-automation teaming. While this level of collaboration holds significant potential for improving performance, it also requires a careful approach to interface design, ensuring users can intuitively guide the AI in unpredictable situations.

Looking ahead

The workshops conducted as part of Validation Exercise #01 and #2 have laid a solid foundation for ASTAIR’s development. By engaging with key stakeholders and addressing real-world operational challenges, the consortium has gained valuable insights that will shape the project’s next phases. As ASTAIR progresses toward developing functional prototypes and refining its algorithms, the lessons learned from these workshops will be instrumental in ensuring the solution is both practical and scalable for widespread use across European airports.

The next stage of validation will involve more in-depth testing, as the project prepares for the Human-in-the-loop Real-Time Simulation planned for later phases, namely Validation Exercise #3, in which all the UCs will be validated. As ASTAIR continues to evolve, the project’s commitment to human-centred design and collaborative decision-making will ensure its success in transforming airport ground operations through cutting-edge AI and automation technologies.

The project is co-funded within the framework of Horizon Europe and brings together the following organisations: Ecole Nationale De L' Aviation Civile (ENAC), Technische Universiteit Delft, Deep Blue, Eurocontrol, Aéroports de Paris.

More about the project