7 AI air traffic programs reshaping Europe and US airspace — which dates are real

Published:

Seven AI and automation programs are reshaping how aircraft are sequenced and routed across Europe and the United States — from the FAA’s $875 million SMART contract to EUROCONTROL‘s AI applications and TopSky-ATC One‘s conflict advisories due in Europe by 2028.

None of them removes a controller from the decision. Only SMART is documented at national scale today; the rest are either narrower, live-but-uncertified, or dated to 2029 and 2030 on timelines no official document confirms as binding.

Ask a traveler what AI is doing to their flight and you’ll get a guess about self-flying jets or a robot at the gate. The reality is narrower and stranger: automation already in place decides which departure slot your aircraft gets and how much spacing it needs on arrival — and in the United States, a local air traffic facility can look at the recommendation and say no.

That veto is the whole story. Air traffic management is one of the last domains where a machine can be genuinely brilliant and still not trusted with the final call.

Europe and the United States have read that problem differently. Washington is scaling a national planning layer first and writing the AI certification rule later. Brussels is drafting the paperwork before the systems go live.

Both approaches are defensible, and neither has published a single number proving the result is better than a human working the scope alone.

Here are the seven programs actually moving, what each one does, and which dates are real.

At a glance

  1. EUROCONTROL AI applications
  2. FAA SMART
  3. SESAR European ATM modernization
  4. HAVEN
  5. NEURON
  6. TopSky-ATC One
  7. TopSky-UAS
Superdeals to Asia preview

Superdeals cover departures from North America, Europe and Australia.

Flight deals most people never see

Air Traveler Club monitors thousands of routes for pricing anomalies that traditional search engines miss. With ATC Superdeals members save up to 80% on each flight vs. regular fares.

Too good to be true? Check it out yourself. It's free.

Newsletter Form

Delivered weekly. Unsubscribe anytime.

The seven projects

1. EUROCONTROL AI applications

AI-based applications are already running inside EUROCONTROL’s network operations.

This is the part most coverage skips: the AI is not coming, it’s already in the pipeline. Turnaround time forecasting, runway capacity management, performance prediction and network-wide flight planning are being shaped by models that have never had a public per-application certification list attached to them. What that means on the ground is quieter than the headlines suggest — slot advice gets sharper, delay knock-on gets predicted slightly earlier, and a controller or flow manager still signs off on the result.

The systems are decision support in the strictest sense, and the public record describes what they do rather than what they deliver.

2. FAA SMART

A 12-year, $875 million contract with Air Space Intelligence underpins SMART, the FAA’s national airspace planning layer.

Roughly 200 data streams — schedules, weather, airport capacity, airspace conditions and operational constraints — feed the system, which turns them into forecasts of traffic flow and flags for conflicts before departure, according to the Federal Aviation Administration. Limited deployment began in September 2026 in Washington, D.C.-area airspace, with expansion described but no binding nationwide completion date established.

Separation responsibility stays with controllers, and local facility leadership can decline the scheduling recommendations. Two hundred inputs, zero final authority over an aircraft — that gap is the entire safety argument in one sentence.

3. SESAR European ATM modernization

59.7% more en-route capacity is the headline figure SESAR projects against a 2012 baseline — a projection, not a measurement.

Europe’s model is distributed by design: shared flight data, trajectory-based operations, dynamic airspace, human-machine teaming, and automation spread across member states, ANSPs and industry partners rather than one national contract. The Master Plan sets deployment objectives running through 2035, which functions as a strategic roadmap, not a procurement commitment.

That distinction matters more than it sounds. A roadmap can be revised without penalty; a signed contract cannot.

4. HAVEN

2030 onward is the stated horizon for HAVEN, a SESAR-backed platform that virtualizes en-route and terminal air traffic services using cloud, automation and AI.

Virtualization is the sleeper here. If controller functions can be delivered from a remote data centre rather than a physical control room, staffing geography stops mattering as much — which is exactly where European ANSPs face their hardest recruiting problem.

The catch is that no independent source confirms 2030 as a fixed service-entry date rather than an aspiration. Treat it as a design horizon until a regulator says otherwise.

5. NEURON

Through 2029, the INDRA-coordinated NEURON project is targeting sustainable taxiing, surface-movement awareness and digital tower visual intelligence.

Of the seven, this is the one closest to a passenger’s actual experience — aircraft moving on the apron, the view from a remote tower, the queue at the gate. Automated taxi routing is routinely described as a fuel and emissions win. No measured fuel or emissions results for NEURON appear in the material reviewed here, so the environmental case remains a design assumption rather than a demonstrated outcome.

That’s not a knock on the engineering; it’s a statement about what has been published.

6. TopSky-ATC One

A 2028 European operational target attaches to Thales‘s TopSky-ATC One, whose core function is AI-driven conflict-resolution advisories for controllers.

This is the deepest incursion into safety-critical decision-making on the list, and Europe’s rules reflect it. The EU AI Act, Regulation (EU) 2024/1689, sets the legal frame, while aviation-specific trustworthiness work runs through the European Union Aviation Safety Agency‘s rulemaking task RMT.0742 and NPA 2025-07 — covering defined operational scenarios, human factors, user interaction and conditions of use. What EASA has published is proposed guidance, not a finished ATM certification rule.

Compliance work to test applications against those expectations is precisely the unglamorous effort that decides whether 2028 is a launch date or a press date.

7. TopSky-UAS

Several countries — among them the United States, Finland and Saudi Arabia — are running TopSky-UAS, Thales’s digital machine-to-machine traffic management system for drones.

Drone integration is the least contested AI surface in aviation because almost nobody is aboard the aircraft. Machine-to-machine deconfliction can be standardized, patched and iterated in ways that a conflict advisory sitting in front of a human controller cannot. It’s also where the gaps between jurisdictions show most plainly: operating approvals vary by country and aren’t itemized in the material reviewed here.

The seven projects, side by side

Verified AI and automation projects in airport and airspace operations
Project Lead entity Region Core function Target operational date Certification or commitment status
FAA SMART Federal Aviation Administration; Air Space Intelligence United States Predictive national-airspace planning using schedules, weather, capacity, airspace conditions, and operational constraints Limited deployment began September 2026; wider completion date unconfirmed Operational deployment confirmed; nationwide completion and AI certification status unconfirmed
SESAR European ATM modernization SESAR Joint Undertaking and European ATM stakeholders Europe Trajectory-based operations, shared data, dynamic airspace, automation, and human-machine teaming Strategic objectives through 2035 Strategic roadmap; not a single binding procurement or certification commitment
HAVEN SESAR-backed project Europe Virtualized air-traffic services using cloud, automation, and AI for en-route and terminal operations 2030 onward, as stated in the brief; independently firm date unconfirmed Aspirational or planned; binding status unconfirmed
NEURON INDRA coordinator and project partners Europe Sustainable taxiing, surface-movement awareness, and digital-tower visual intelligence Through 2029, as stated in the brief; independently firm date unconfirmed Research-project target; binding certification status unconfirmed
TopSky-ATC One Thales Europe AI-supported conflict-resolution advisories for controllers 2028, as stated in the brief; independently firm date unconfirmed Planned vendor deployment; regulatory commitment unconfirmed
EUROCONTROL AI applications EUROCONTROL Europe Network flight planning, performance prediction, runway-capacity management, and turnaround forecasting Already in use; exact deployment dates for individual applications unconfirmed Operational applications reported by EUROCONTROL; common certification status not established
TopSky-UAS Thales United States, Finland, Saudi Arabia, and other countries Digital machine-to-machine traffic management for drones Operational in several countries; individual launch dates unconfirmed Operational deployment reported; jurisdiction-specific approvals vary and are not itemized in the supplied material
Source: Federal Aviation Administration (FAA); SESAR; European Union Aviation Safety Agency (EASA)

Notice what the table can’t tell you. There is no column for measured performance, because no controlled, at-scale comparison of AI-integrated airspace operations against legacy ones turned up in any independent source.

The bodies capable of producing one — the FAA’s national performance metrics and EUROCONTROL’s Performance Review Commission — haven’t published it. Until they do, every efficiency claim in this space is a forecast.

How we picked

Each entry traces to official FAA, EASA, SESAR or program material — nothing earned a place on the strength of a vendor announcement alone. Where a date came from the brief rather than a confirmed document, the entry says so. Airport-side automation frequently lumped into this conversation — automated baggage systems, biometric gates, predictive-maintenance AI, autonomous apron vehicles — was left out because no sufficiently specific deployment record was verified; September 2026 remains the most recent confirmed deployment milestone in the material reviewed.

What this means for you

The regulatory split has practical consequences for anyone booking across both systems. European deployments will likely arrive later but with a documented compliance trail behind them; the U.S. model moves faster and puts its safety bet on controllers keeping the final call, with agency review layered on top. Neither is obviously safer. They’re different bets about where to put the brake.

Expect automation to touch your trip in ways you won’t see — slot allocation, departure sequencing, surface routing — long before anything as visible as a biometric gate. Don’t assume uniformity. The same airline can fly you through a heavily automated network in one direction and a largely manual one on the return.

One open question deserves flagging: nothing in the reviewed material names an Asia-Pacific AI or air-traffic-management deployment specific enough to compare with these seven. For travelers on trans-Pacific routings, that’s an information gap, not an absence of activity. The human side is less documented still: the reviewed material covers governance and deployment planning, not how controllers actually work with these advisories over time.

ATC Intelligence is the research division of Air Traveler Club. Backed by 15 years in Asia-Pacific aviation, we don't just report on the regional market; we live and work in it. By pairing AI-driven data with strict human fact-checking, we provide actionable, trustworthy journalism designed to make your trips to Asia smarter and more affordable.

Follow our daily coverage on Google News, Google Discover, and social media.

Key terms

SMART
SMART is the Federal Aviation Administration’s national airspace planning system, built to forecast traffic flow and flag conflicts before departure. It runs on a 12-year, $875 million FAA contract with Air Space Intelligence and combines roughly 200 data streams, including weather, airport capacity, airline schedules, airspace conditions and operational constraints. For all its inputs, it never gains final authority over an aircraft — the gap that defines the U.S. approach to AI in air traffic management.
SESAR
SESAR is Europe’s air traffic management modernization program, run through a joint undertaking with member states, air navigation service providers and industry partners. Its Master Plan sets deployment objectives through 2035 and projects gains against a 2012 baseline, including 59.7% more en-route capacity and 19% better punctuality. Unlike a national procurement, it functions as a strategic roadmap that can be revised without penalty, which is why its targets are not binding commitments.
ANSP
An air navigation service provider, or ANSP, is the organization responsible for providing air traffic services within a defined block of airspace. In Europe, SESAR’s distributed model spreads automation across member states, ANSPs and industry partners rather than concentrating it in a single national contract. That structure means automation arrives through many separate providers rather than one national system, the sharpest contrast with the FAA’s single SMART contract.
EU AI Act
The EU AI Act, Regulation (EU) 2024/1689, is the European Union’s legal framework for artificial intelligence. For aviation, its requirements are being developed through EASA’s rulemaking task RMT.0742 and NPA 2025-07, which address operational scenarios, human factors, user interaction, conditions of use and limitations. What exists so far is proposed guidance rather than a finished air traffic management certification rule, so the 2028 target for TopSky-ATC One sits ahead of the standard that will govern it.
Trajectory-based operations
Trajectory-based operations is an air traffic management concept in which flights are managed as continuous, shared trajectories rather than as a series of point-to-point instructions. SESAR lists it among the building blocks of its distributed model, alongside shared flight data, dynamic airspace and human-machine teaming. Because it depends on sharing trajectory data across borders and providers, it spreads automation across many stakeholders instead of concentrating it in one national system.

Questions? Answers.

Why can’t AI replace air traffic controllers?

Under the FAA’s SMART program, separation responsibility stays with controllers. The system turns multiple data streams into forecasts and scheduling recommendations, which FAA personnel review — and local facility leadership can decline them. The contract buys a planning layer, not autonomous control of aircraft.

Are the 2028, 2030, and 2029 target dates for TopSky-ATC One, HAVEN, and NEURON firm commitments?

No. Those dates are planned or aspirational targets rather than confirmed regulatory or procurement commitments. SESAR’s Master Plan sets objectives through 2035 as a strategic roadmap, and no independent source in the reviewed material establishes a binding operational start for any of the three.

Has any independent study compared AI-integrated airspace operations against legacy operations at scale?

Not in the material reviewed. The published figures are program projections rather than independent evaluations of deployed AI. A defensible comparison would need to draw on the FAA’s national performance metrics and EUROCONTROL’s Performance Review Commission, and no such AI-versus-legacy study was found.

What does Europe’s regulatory framework require of AI systems used in aviation?

The EU AI Act, Regulation (EU) 2024/1689, provides the legal frame, with aviation-specific requirements being developed through EASA’s rulemaking task RMT.0742 and NPA 2025-07. Those proposals cover defined operational scenarios, human factors, user interaction, conditions of use and limitations. EASA has published proposed guidance, not a final certification rule for air traffic management.