FAA Announces AI Models To Fix Safety And Congestion Issues

The Federal Aviation Administration (FAA) has announced several new AI tools to address ongoing problems with flight safety and congestion. As part of a broader push to modernize America’s aging air traffic systems, the FAA has introduced AI models that will synthesize data from distinct databases to analyze safety issues and congestion.

One AI model, called Foundry, made by defense contractor Palantir, will look for identifiable patterns in airport safety issues by analyzing data from various sources. The first year of development is estimated to cost $4 million.

Foundry will review data to identify repeatable behaviors or hotspots that increase statistical risk, but it will not predict individual instances. The FAA has also announced a separate AI model, created by Boston defense company Air Space Intelligence (ASI), to ease traffic congestion by predicting bottlenecks and delays.

The ASI model, called Strategic Management of Airspace, Routes and Trajectories (SMART), will be the new technological backbone of the FAA’s Air Traffic Control System Command Center. The SMART system will cost $875 million and is scheduled for rollout in September.

The FAA plans to stage the rollout, initially using the AI system only on traffic above 24,000 feet. The introduction of AI systems to air traffic control has raised questions about whether they need more testing before being introduced.