New AI path-planning algorithm for air traffic control computes routes in 3.69ms

A new paper from Zou et al. tackles a long-standing gap in air traffic management: path-planning algorithms that work with controllers rather than against them. The team’s “Solution Space Path Planning” (SSPP) algorithm is built from the ground up to be interpretable and flexible, using a solution-space display paradigm that shows all feasible safe actions simultaneously. It incorporates three intent-based conflict detection methods—distance-based, time-interval-based, and zone-based—to identify conflict-free paths efficiently. Two variants were tested: SSPPV (vertex-based search) and SSPPE (edge-based search), with SSPPV paired with zone-based detection emerging as the clear winner.

In tests using real operational data from the Delta sector of the Maastricht Upper Area Control Centre (MUAC), SSPPV computed safe paths in an average of just 3.69 milliseconds on a 5 nautical mile grid. The algorithm prioritizes operational constraints like separation standards, maneuverability limits, and waypoint minimization, making it a practical decision-support tool. This represents a significant step toward closing the gap between algorithmic research and real-world controller needs, potentially enabling safer, more efficient en-route traffic management without sacrificing human oversight.