Archer says ZEE AI model predicts airport surface aircraft trajectories minutes ahead
ACHR•ZEE model aims to predict ground-movement risk
Archer Aviation announced a technical breakthrough for its ZEE aviation foundation model, predicting airport-surface aircraft trajectories minutes ahead in real time.
The system targets runway-related risk, which accounts for 30% to 40% of global aviation accidents, aiming to add a predictive safety layer for controllers.
ZEE uses conditional flow matching to model multiple possible ground-movement futures, improving on single-path prediction approaches.
A vision transformer trained on high-resolution satellite imagery maps runways, taxiways, and aprons to anchor predictions in physical airport layouts.
Testing has begun at Hawthorne Airport, with larger-scale trials underway, as Archer plans work with government agencies toward operational validation.
