As the ORCI solution progressed towards higher levels of technological maturity, it became essential to evaluate its performance in a realistic operational environment. This validation step was designed to assess not only the technical capabilities of the tool, but also its potential to support air traffic controllers during arrival operations.
The ORCI project designed a comprehensive validation campaign to assess how aircraft spacing predictions could support controllers during arrival sequencing. Rather than evaluating the machine learning models in isolation, the project focused on understanding how predictive information could be integrated into operational decision-making.
To achieve this, ORCI selected two representative operational environments covering different arrival management concepts currently used in Europe. These scenarios allowed the project to evaluate the solution across different traffic situations and operational procedures while ensuring that the predictive models were tested under realistic conditions.
The validation activities combined advanced simulation with Human-in-the-Loop (HITL) exercises, where professional air traffic controllers interacted with the ORCI Decision Support Tool while managing simulated arrival traffic. This approach made it possible to evaluate not only the accuracy of the predictions, but also their usability, operational relevance, and potential impact on controller decision-making.
Distributed Validation Environment
To recreate these operational conditions, ORCI employed a fully distributed hybrid simulation environment. The simulations were performed using RAMS Plus, an advanced air traffic management simulation platform capable of reproducing realistic aircraft behaviour and controller operations developed by ISA Software.
The simulation engine was hosted in Paris, while controllers participated from dedicated working positions at the Barcelona and Lisbon Area Control Centres (ACCs) using customised Human-Machine Interfaces (HMIs) that replicated their local operational environment. At the same time, the ORCI AI-based Decision Support Tool, hosted in Madrid, generated aircraft spacing predictions in real time using the machine learning models developed within the project.
This distributed architecture allowed all system components—including the simulation platform, controller interfaces, and AI services—to communicate seamlessly throughout each validation exercise, closely replicating how such a solution could operate in a future deployment.
Validation Exercises
To evaluate the ORCI Decision Support Tool under different operational conditions, the project carried out two validation exercises representing two distinct RNAV arrival management concepts: Barcelona (Trombone) and Lisbon (Point Merge).
For both case studies, controllers managed realistic arrival traffic in two scenarios: a reference scenario, using standard operational procedures, and an ORCI-assisted scenario, where real-time spacing predictions were available through the Decision Support Tool. This approach enabled a direct comparison of operational performance with and without ORCI support.
The validation campaign involved 16 active air traffic controllers—eight from ENAIRE and eight from NAV Portugal. In addition to simulation performance data, the evaluation incorporated controller feedback through debriefing sessions and post-simulation questionnaires, providing both quantitative and qualitative insights into the operational value of the solution.
| Validation Exercise | Barcelona RWY 24R | Lisbon RWY 02 |
|---|---|---|
| Arrival concept | Trombone | Point Merge System |
| Runway configuration | Segregated operations | Mixed-mode operations |
| Validation scenarios | Reference + ORCI | Reference + ORCI |
| Participating ATCOs | 8 | 8 |
| Main objective | Assess ORCI performance during Trombone operations | Assess ORCI performance in a Point Merge environment |
Validation platform
•Simulation with connected ML-based predictive model
•Continuously predicts spacing for identified arrival pairs at ILS interception/PMS merge point
•Refresh rate target < 5 seconds
•Transparent AI approach to maintain ATCO trust
•Aircraft turn dynamics/momentum
•Wind & compression effects
Barcelona Demo
Lisbon Demo
KEY RESULTS
The validation campaigns provided valuable insights into the potential benefits of the ORCI Decision Support Tool in both Barcelona and Lisbon. Compared with the baseline scenarios, the use of ORCI improved spacing accuracy by 57% in Barcelona and 39% in Lisbon, while maintaining safety performance with no increase in separation-related issues. The results also indicated improvements in runway arrival throughput and operational efficiency, with reductions in the nautical miles flown during the arrival sequence.
The positive impact extended beyond operational performance. The results showed reduced fuel consumption, with estimated savings of 48 kg of fuel per arrival in Barcelona and 21 kg in Lisbon, while controller feedback indicated high acceptance of the tool and a reduction in workload and stress. Together, these results highlight the potential of predictive aircraft spacing information to support more accurate, efficient, and sustainable arrival operations without compromising safety.
Follow ORCI on LinkedIn to stay up to date with the latest news.