SMARTS wraps up - paving the way for smarter airspace

As the SMARTS project concludes after 2.5 years of research and collaboration, it leaves behind more than algorithms and prototypes. SMARTS has outlined a pathway to making Europe's airspace more flexible, predictable, and resilient - capable of adapting to evolving traffic patterns and operational pressures while reducing delay measures. The project aimed to prepare the airspace for changing demand, even under challenging conditions like weather disruptions or traffic peaks.

A project built on collaboration and innovation

Led by CRIDA and powered by a consortium of operational, academic and research partners (ENAIRE, EUROCONTROL, NATS, ENAC, DLR and Lancaster University), SMARTS combined artificial intelligence, mathematical optimisation and operational experience to rethink how sectors and configurations are created.

The project explored how to:

• Understand traffic demand with machine learning
• Build airspace "building blocks" that adapt to traffic
• Design sector shapes aligned with real flows
• Automatically propose the best configuration plan for each day

Figure 1 SMARTS partners

These elements were integrated into tools and methods that support airspace preparation from months up to a few hours before operations. SMARTS delivered a complete set of building blocksfor future demand capacity balancing (DCB): flow clustering, machine learning (ML) based sector profile forecasting, Basic Volumes, sector design methods, and airspace configuration optimization - integrated through PUZZLE and R-NEST, which also provide the user interface. These were tested and validated in two major exercises.

SMARTS solution

The SMARTS solution strengthens the resilience and adaptability of sectors and sector configuration planning through advanced optimisation and artificial intelligence (AI) techniques.

At its core, SMARTS designs"smart" sectors and improved sector configurations that dynamically allocate air traffic demand while maintaining balanced air traffic controller (ATCO) workload. By optimising sector design and shapes, the concept enables a more efficient and capacity-driven use of airspace. As a result, the number of required demand-capacity measures can be reduced, contributing to higher smoother operations. AI and machine learning (ML) techniques support this process by providing insight into the expected environment and identifying optimal sector designs and configurations.

Moreover, the demand analysis sub-systems apply ML techniques to flight plan data and actual flight trajectories. This enables the identification of typical air traffic "flows" and trajectory predictions, which are then used to forecast demand for a given sector configuration.

By the end of the project, the SMARTS solution successfully reached TRL2.

Figure 2 Demand and capacity balancing processes addressed by the SMARTS solution

Two major exercises to test the concept in Madrid ACC

CRIDAran conducted real-time simulations in which flow management positions (FMPs) used the PUZZLE tool to test SMARTS concept in realistic traffic conditions with DLR providing key performance indicators on resiliency. ENAIRE and NATS FMPs assessed new sector shapes, configuration options, and the tool's ability to support decision-making under normal, adverse weather, and high-demand scenarios.

Figure 3 PUZZLE Screenshot showing FMP functions to support sector configuration planning

In these high-complexity environments, FMPs evaluated:

• Operational feasibility of SMART sector shapes and configurations
• Improvement on enhanced situational awareness
• Effectiveness of automated support in solving demand-capacity imbalances
• Early performance impacts on capacity, fuel efficiency, cost efficiency, and punctuality

Feedback was encouraging: sectors aligned with traffic flows, the tool provided clear decision support, and SMARTS configurations reduced overloads and lowered the need for demand measures.

Figure 4 CRIDA Real-Time Simulations carried out in June 2025

EUROCONTROL tested the enhanced R-NEST algorithm, which automatically generates sectorisations and configuration plans. Using real flight plans and 2030 traffic projections.

Figure 5 R-NEST Screenshot showing sector design and configuration functions

Results showed:

• Significant reductions in overload
• Fewer control position hours
• Balanced workload across the ACC
• Smoother, more efficient opening schemes

In particular, the30-sector solution for Madrid ACC, offered an excellent balance of efficiency, stability, and operational feasibility.

Figure 6 EUROCONTROL real-Time Simulations carried out in June 2025

SMARTS models and algorithms

1. Flow clustering model. The flow clustering model automatically identifies major 3D traffic flow structures from flight data, providing an accurate and data-driven representation of traffic patterns. This supports improved demand prediction and optimised sector configuration planning.

Figure 7 SMARTSFlow Clustering model architecture developed by DLR

1. Sector profile prediction model. The sector profile model enhances tactical planning by providing data-driven traffic demand forecasts up to 8 hours in advance. By integrating flight intentions, weather impact, and sector capacity constraints, it supports more resilient and adaptive sector configuration decisions. A dedicated local resilience KPI further strengthens decision-making by quantifying the robustness of configuration plans.

Figure 8 SMARTSSector Profile Prediction model developed by DLR

1. Basic volumes design model. The basic volumes design model introduces advanced algorithms for dynamic sector design, aimed at optimising airspace blocks (ABs) and shareable airspace blocks (SABs). By combining simulated annealing techniques with complexity-based performance metrics, the model generates optimised airspace building blocks that can flexibly adapt to evolving traffic patterns and operational conditions.

Figure 9 Identification of Airspace Blocks (Green) and Shareable Airspace Blocks (Blue) based on traffic complexity assessment made by the Basic Volume model developed by ENAC

1. Sector design model. The model represents airspace as a graph of basic volumes, with edges indicating adjacency and weights reflecting traffic and coordination workload. It groups these volumes into operational sectors while ensuring connectivity, convexity, and minimizing inter-sector coordination. The algorithm efficiently generates optimal 2D sectors within minutes and near-optimal 3D sectors within an hour. It integrates traffic flows and airspace blocks to produce sector catalogues ready for practical deployment. Overall, the model supports safer, more flexible, and scalable airspace management.

Figure 10 Examples of SMARTS Sector Design made by SMARTS Sector Design Model developed by University of Lancaster. The model applies optimisation and operational criteria to combine SMARTS Basic Volumes.

1. Sector configuration model. The model is a computational framework for dynamic airspace configuration that optimizes sector workloads and adapts to traffic fluctuations. It combines machine learning traffic predictions with intelligent sector design to generate efficient, balanced configurations. Advanced optimisation algorithms enable timely, operationally feasible planning. A user interface and API allow real-time monitoring, analysis, and implementation of optimal sector configurations.

Figure 11 Example of SMARTS Sector Configuration Plan made by SMARTS Sector Configuration Model developed by University of Lancaster. The figure represents a proposal in which four sector configurations are deployed during the dayshift. On the right side, the graphical representation of one of the configuration in respect to the main traffic flows is represented.

What we learned

SMARTS confirmed that smarter, data-driven airspace management brings clear benefits across multiple areas:

• Capacity: Better alignment of sector boundaries with actual traffic flows
• Operational efficiency: Fewer overloads and reduced need for traffic regulations, meaning smoother days for controllers
• Cost efficiency: Fewer sector openings and configuration changes
• Punctuality & flight efficiency: Reduced demand measures lead to fewer delays and more direct trajectories
• Resilience: Configurations better cope with disruptions, boosting overall system robustness
• Human performance: Easy-to-use tools provide FMPs with clearer, more actionable insights for faster, informed decisions

Both PUZZLE and R-NEST received strong feedback for usability and operational value. Across all exercises, SMARTS demonstrated that algorithm-supported, data-driven airspace management works and reached TRL2, confirming its operational feasibility and technical robustness.

Figure 12 Summary of the main benefits provided by SMARTS Solution.

Looking ahead

SMARTS has demonstrated that automation and AI can play a meaningful role in demand capacity balancing. Several of its automation components already reach automation levels 2-4, making them promising enablers for future SESAR industrial research projects.

The project highlighted multiple areas for further development and research:

• Refining the definition and operational use of shareable airspace blocks (SABs) across Europe.
• Improving demand prediction models, particularly to enhance tactical accuracy.
• Extending validation to additional ACCs, cross-border contexts, and seasonal variations.
• Strengthening resilience metrics to better prioritise capacity-related robustness.
• Expanding sector catalogues to cover all seasonal patterns.

Overall, SMARTS outcomes were strong and provide a solid foundation for further SESAR research, while offering actionable opportunities to improve operational efficiency, predictability, and resilience in airspace management.

A solid step toward the future of ATM

SMARTS concludes with a clear message: dynamic, intelligent airspace is possible-and it works. By delivering proven benefits in capacity, efficiency, and resilience, the project provides a solid foundation for future SESAR developments and sets the stage for the next generation of airspace management.

More than a set of tools, SMARTS leaves behind a structured, validated, and forward-looking concept for dynamic airspace configuration, offering measurable operational benefits. Thanks to the dedication of the consortium and the operational experts involved in the validation campaigns, SMARTS stands as a key stepping stone toward a smarter, more adaptable, and sustainable ATM network.

More about the project