Auto-Suggest Capability via Machine Learning in SMART NAS

Period of Performance: 06/17/2015 - 12/17/2015

$122K

Phase 1 SBIR

Recipient Firm

The Innovation Laboratory, Inc.
2360 SW Chelmsford Avenue
Portland, OR 97201
Firm POC
Principal Investigator

Abstract

We build machine learning capabilities that enables the Shadow Mode Assessment using Realistic Technologies for the NAS (SMART NAS) system to synthesize, optimize, and "auto-suggest" optimized Traffic Management Initiatives (TMIs). Multi Level Multi View (MLMV) machine learning is used to identify similar historical situations (days, scenarios, or airport conditions) in the NAS. TMIs used in historically similar situations are locally modified to optimize the parameters of the TMI to be used in the current day situation. SMART NAS is used to evaluate TMIs and to present fast time simulations to the end user to review the TMI and associated performance metrics before implementation.