Skip Navigation

Machine Learning of Multi-Modal Influences on Airport Delays, Phase I

Completed Technology Project

Project Introduction

We build machine learning capabilities that enables improved prediction of off-block times and wheels up times which are critical inputs to NAS stakeholders. NextGen will rely on machine learning techniques utilizing all sources of useful information in order to improve predictive accuracy and reliability of flight operations in the NAS. These predictive capabilities will support real-time optimization of surface operations. We use machine learning to learn from historical data and similar situations in the past in order to optimize the performance of the NAS for the current situation. The proposed Multi-Level, Multi-View (MLMV) machine learning approach takes real-time weather, demand, and other data inputs (including landside data from TSA security line queues and traffic congestion levels on highways), searches through an archived set of historical data, identifies similar situations and NAS controls used in those situations, ranks historical situations according to their effectiveness, estimates a set of Traffic Management Initiatives (TMIs) and other control strategies impacting off-block-times and wheels up times. More »

Anticipated Benefits

Primary U.S. Work Locations and Key Partners

Share this Project

Organizational Responsibility

Project Management

Project Duration

Technology Maturity (TRL)

Target Destinations

Light bulb

Suggest an Edit

Recommend changes and additions to this project record.