Go to Page # Page of 19

Using Big Data and GIS to Model Aviation Fuel Burn

 Gary M. Baker
Description: Traditional Databases vs. Hadoop:Sports Car-refined,has a lot of features,accelerates very fast,pricey,expensive to maintain.Freight Train-rough,missing a lot of 'luxury',slow to accelerate,carries almost anything,moves a lot og stuff very efficently.Input Flight Data: Flight Track -National Airspace System (NAS):radar based,latitude,longitude,altitude,time.Other parts of the world:airways,waypoints,great circle.Altitude Profile-Altitude varies throughout the flight affecting fuel burn.Fuel Burn Model:Model outputs many segments for each flight,Fuel burn is uniform over segment,Average of over 100 segments per flight.Route Variations:One year of flights from LAX to BOS-Winds,Weather,Traffic,Restrictions,Procedures.This is only one of over 850,000 origin destination pairs!Altitude Variations:Winds,Weather, Traffic,Restrictions,Procedures,Load,Equipment.Objective:Intersect fuel burn segments with 36 km Equal-Area Scalable Earth Grid,Determine total fuel burn for each grid cell.Sounds Like a Big Data Problem:~ 30 million flights globally per year,average of over 100 segments per flight,multiple analysis years,different scenarios.We could use traditional statistical approaches to model flight tracks and altitudes but that’s not the big data way!Steps:Preprocess data - reformat, subset, project, Write and debug program locally (Java),Adapt program to MapReduce model and Hadoop environment,Overcome cloud learning curve-connect, execute, debug, automate, etc.Upload data to the cloud,Execute and retrieve results.
Views: 1746
Domain: Energy
Category: Transportation
Using Big Data and GIS
to Model Aviation Fuel Burn
Gary M. Baker
USDOT Volpe Center

2015 Transportation DataPalooza
June 17, 2015

The National Transportation Systems Center
Advancing transportation innovation for the public good

U.S. Department of Transportation
Office of the Secretary of Transportation
John A. Volpe National Transportation Systems Center

Presentation Outline

Big Da ... See more

Recent Presentations

01 October, 2019
30 September, 2019
29 September, 2019