AI tool to improve arrival sequencing and reduces controller workload in complex terminal airspace

The SESAR JU ORCI project has completed two successful human-in-the-loop validation campaigns demonstrating how its AI-powered predictive spacing tool can support air traffic controllers during arrival sequencing in complex terminal manoeuvring areas (TMAs). Conducted with ENAIRE and NAV Portugal, the simulations involved 16 experienced air traffic controllers working realistic, high-density traffic scenarios at Barcelona-El Prat and Lisbon airports. The results provide strong evidence of operational relevance and scalability across different approach environments.

The validation activities were designed to assess how the ORCI predictive tool can support controllers during the critical phase of arrival sequencing by providing real-time predictions of the spacing that would result from a given vectoring instruction. This predictive capability allows controllers to better anticipate the outcome of their tactical decisions, contributing to more stable and predictable arrival flows under high traffic demand.

The validation campaigns included baseline scenarios using current operational tools and equivalent scenarios with ORCI support, both designed with comparable traffic demand and complexity. All scenarios featured very high traffic loads, exceeding nominal operational capacity, in order to evaluate ORCI performance under demanding conditions. Performance metrics were collected through simulation outputs, post-simulation questionnaires, and air traffic controller debriefings.

In Barcelona, using a parallel trombone approach in a segregated arrivals-only runway configuration, controllers were instructed to achieve target spacing at localiser interception based on RECAT wake turbulence separation criteria. In Lisbon, using a Point Merge System operating in mixed-mode traffic on a single runway, controllers aimed to achieve target spacing at the Point Merge System merge point, depending on whether a departure needed to be inserted between consecutive arrivals, with spacing objectives coordinated in advance. In both environments, ORCI demonstrated flexibility in supporting a wide range of spacing objectives.

Across both validation campaigns, results show that ORCI significantly improves spacing stability compared to baseline operations. Controllers were able to achieve target spacing more consistently, reducing dispersion in spacing outcomes and limiting the occurrence of non-conservative situations. In Barcelona, the mean absolute spacing error was reduced from 0.86 NM in baseline operations to 0.45 NM with ORCI support, representing a 47% reduction. In Lisbon, the mean absolute spacing error decreased from 0.53 NM to 0.30 NM, corresponding to a 43% reduction.

Furthermore, controllers noted that the ORCI tool provides reliable predictive information that reduces cognitive workload by limiting the need for continuous mental estimation of spacing and turn timing. This qualitative feedback was supported by post-validation workload assessments, which showed that perceived controller workload was reduced by almost half when using ORCI, while fully preserving air traffic controllers' responsibility for spacing and timing decisions.

Feedback gathered during validation debriefings highlighted the high realism of the distributed simulation platform, both in its visual representation and in traffic behaviour, with some participants noting that it could even be suitable for training purposes. Several controllers stated that they would like to have the ORCI AI tool available at their working position "tomorrow". ORCI was also seen as particularly beneficial during the warm-up period at the start of a shift, typically the first 10-15 minutes, when controllers are still building situational awareness of traffic characteristics.

In addition, controllers reported that ORCI helps reduce "chain reactions", where an initial spacing deviation would otherwise require multiple corrective actions through speed and vectoring adjustments, potentially leading to further inefficiencies in the inbound stream. By reducing the need for continuous mental spacing calculations, ORCI frees cognitive resources and allows controllers to focus more effectively on managing complex or non-standard situations.

The successful completion of these validation campaigns represents an important step towards the operational maturity of the ORCI concept. The results demonstrate its potential to support more predictable arrival management, reduce controller workload (LPPT: 43%, LEBL: 43%), and contribute to more efficient use of runway throughput, while fully preserving controllers' final decision-making role.

More about ORCI