SafeTeam’s Approach to Managing Air Traffic Complexity

Pablo Hernández

2024-06-12 10:21:41
Reading Time: 5 minutes

The introduction of AI-based tools into ATC systems is set to significantly transform human-system interactions. Recognizing this, the SafeTeam project was set out to develop a comprehensive framework to address Human Factors (HF) and improve human-machine collaboration. The goal of the SafeTeam framework is to ensure that human operators can seamlessly coordinate with these advanced digital assistants, addressing essential challenges and constraints at an early stage of the design cycle.

Tailored for researchers and practitioners with limited experience in Human Factors and Human-Automation Interaction (HAI), the SafeTeam project’s methodological approach offers valuable guidelines and actionable insights. It assists in evaluating and monitoring system performance, emphasizing safety and resilience to ensure seamless cooperation between humans and machines. Throughout the project, the framework will be applied in various operational use cases, such as digital assistants for enroute air traffic management and aircraft cockpit HMIs for predicting and warning pilots about unstable approaches. Additionally, it will support the development of regulatory and certification requirements for automation tools to meet market needs and gain societal awareness.

In this blog post, we dive the first operational use case, led and currently under development by DataBeacon. The study aims to evaluate the use of digital assistant in air traffic control (ATC) by anticipating complex traffic scenarios, forecasting airspace sector workload and allowing more efficient and environmentally-friendly routing—all while increasing safety.

In a previous work package (WP 3.1), SafeTeam conducted extensive surveys among air traffic controllers (ATCOs) to gather valuable feedback on the factors influencing air traffic complexity and workload. These surveys revealed critical insights into the distinction between conflicting and non-conflicting traffic and their impact on airspace complexity. Traditionally, airspace capacity has been determined primarily by the occupancy or density of traffic, closely linked to the configuration of airspace, sectors, flows, and airways. However, with the shift towards a free route airspace—where flights can move from any point to any point without the constraints of predefined airways—the dynamics of air traffic management are evolving.

The findings from WP 3.1 highlighted the necessity of incorporating additional Key Performance Indicators (KPIs) that measure traffic patterns requiring higher ATCO involvement. These KPIs, such as Potential Conflict Detection (PCD) vs Non-Conflicting Traffic (NCT), offer a more nuanced understanding of air traffic complexity beyond mere traffic density. By considering these factors, complexity management tools can provide a more balanced workload across sectors, enhancing both efficiency and safety. This evolution in complexity management is essential as we move towards a more flexible and efficient airspace system, capable of adapting to the new demand on air traffic while maintaining high standards of safety and operational efficiency.

Building on these insights, the SafeTeam project has developed key certain performance indicators (KPIs) to better understand air traffic complexity. These KPIs include Potential Conflict Detection (PCD), Non-Conflicting Traffic (NCT) at first sight, Occupancy Count (OCC), Sector Capacity, as a dimensional way of measuring the current situation, and Aircraft Maneuvers, which will certainly show a high degree of correlation with the previous indicators.

  • Potential Conflict Detection (PCD): A high number of PCDs indicate areas and time frames where aircraft trajectories are more likely to produce loss of standard separation, requiring an increase in controller intervention and augmenting the complexity.

  • Non-Conflicting Traffic (NCT): A high number of NCTs suggest patterns on air traffic that do not require much involvement from the ATCO, referring to over-flying traffic that should not require deviations in the trajectory or further monitoring. Overall, high NCT would correlate with lower airspace complexity, and thus a decrease in the controller workload.

  • Occupancy Count (OCC): Higher occupancy counts reflect greater traffic density, contributing to increased complexity due to the higher volume of aircraft to manage.

  • Sector Capacity Utilization: This metric compares actual occupancy to the sector’s capacity, indicating how close operations are to their maximum safe limits. High utilization rates increase complexity as sectors approach their capacity.

  • Aircraft Maneuvers: Frequent maneuvers indicate active piloting and control efforts, adding to the complexity as controllers must manage dynamic changes in aircraft positions and paths.

Moreover, the team is working on development of interactive interfaces to help controllers work more effectively and support their decisions based on data and complexity base KPIs. Although the project is still in its development phase, three distinct views are under consideration: a traffic view that offers real-time or replay traffic in ATCO style, a configuration comparator with pre-defined constraints on valid sectorizations, and a sector playground where the KPIs can be compared by time frames and for any sector.

  • Traffic Map: This view allows users to replay and analyze traffic data, correlating actual trajectories with PCDs and NCTs. It supports validation of complexity metrics by providing a visual and interactive means to explore specific traffic scenarios, helping controllers understand areas of low or high complexity.

  • Configuration Comparator: This tool enables the comparison of different sector configurations to identify the most effective airspace allocation that balances the complexity across sectors. This view helps minimize peak complexity and variance, facilitating smoother and safer air traffic operations.

  • Sector Playground: Designed for exploring and comparing metrics and complexity across various sectors. In this view, users can customize the display and analyze sector-specific data without constraints on the allowed configuration of the airspace.

The dashboard purpose is not just to provide a visualization tool but to foster a collaborative environment where human expertise and the digital assistant complement each other. The future real-time feedback and detailed analyses provided by these sort of tools will empower controllers to make informed decisions, ultimately enhancing both efficiency and safety in air traffic management.

While this solution offers an innovative approach as a decision support tool on complexity management, it also opens up exciting opportunities for further enhancement. One promising direction is integrating predictive AI models to forecast potential conflicts and non-conflictive traffic, enabling the dashboard to not only display past and current traffic but also predict future scenarios. That is, trying to predict conflicts and non-conflictive traffic before it has actually happened. Another interesting feature could be to find the best configuration automatically, it is said, suggest to the supervisor ATCO what would be the optimal sectorization layout of the airspace that balances the complexity for a given set of resources.

Another great improvement would be to gather data from other countries and extend the scope of the dashboard so it can be used for different airspaces, besides processing as much data as possible to cover a longer time period.

Finally, as part of the next steps and milestones in the SafeTeam project, we are committed to and aim to providing operationally mature solutions with a high technology readiness (TRL 6). To this end, we have planned dedicated feedback validation workshops and exercises with air traffic controllers (ATCOs). These will focus on collecting feedback on the complexity metrics and interactive interfaces, as well as the implementation and evaluation of the performance metrics defined in the SafeTeam’s framework. By engaging directly with ATCOs, we aim to refine this digital assistant tool based on real-world feedback, ensuring it meets operational needs and enhances the efficiency and safety of air traffic management. Stay tuned for updates on these workshops, our progress in other case studies, advancements in the Human-Factors framework, and our work in certification and regulation issues.

Authors: Pablo Hernández, Eugenio Neira and Ernesto Gregori