Part 2: SMART and the Four-Phase Framework

Key Takeaways

• SMART extends conflict prediction from 15 minutes to two hours
• FAA plans vendor selection by end of May, deployment by September 2026
• Procurement timeline compressed to five months, far faster than typical safety‑critical software
• Political pressure drives speed, raising concerns over verification rigor
• Vendors lack ACAS X‑level verification experience; risk of advisory fatigue

Pulse Analysis

The Strategic Management of Airspace Routing Trajectories (SMART) represents the first layer of a four‑phase roadmap the FAA envisions for automating air‑traffic management. In its Phase 1 incarnation, SMART ingests flight‑schedule, weather and airspace‑state data, runs four‑dimensional modeling, and surfaces conflict‑avoidance recommendations to controllers. By extending the predictive horizon to two hours, the system promises earlier deconfliction, potentially smoothing peak‑hour traffic and reducing downstream delays without stripping human decision‑making authority.

What sets SMART apart is not just its technical ambition but the velocity of its procurement. The FAA has moved from a public vendor competition announcement to an operational target in roughly five months—a timeline that dwarfs the 15‑year development arc of ACAS X and the sluggish progress of the broader NextGen suite, which has consumed $7.5 billion for only 16 % completion. This acceleration stems from political pressure after high‑profile collisions, with senior officials and even private influencers urging rapid safety upgrades. The trade‑off is a compressed verification window, raising doubts about whether the rigorous testing that secured ACAS X’s safety record can be replicated.

The three shortlisted vendors each bring distinct strengths but also gaps. Air Space Intelligence offers a proven AI flight‑path optimizer used by a major carrier, yet its experience is limited to commercial dispatch environments. Palantir’s deep FAA data‑integration via its Foundry platform provides a ready‑made data backbone but lacks a track record of safety‑critical AI validation. Thales contributes decades of ATM heritage, though its U.S. deployment experience is thin. None match the 15‑year verification pedigree of MIT Lincoln Laboratory’s ACAS X program, leaving the FAA to balance speed against assurance. If SMART’s advisories prove accurate, the modest throughput gains could validate fast‑track AI adoption; if not, controllers may ignore the system, relegating it to shelfware and prompting a costly re‑investment in NextGen 2.0 while the underlying staffing and infrastructure challenges remain unresolved.