Air Traffic Control - H Optimize Pushback,Orthe Rate at Which Instructions Used to Solve Aproblem) to on Algorithms (Aset Ofstep-By-Step Range Ofconditions

Home , Air traffic control

“ from thereal workld. that incorporate data velop better algorithms mental research to de it allbegins withfunda- people interact with.And of systems that billionsof prove theperformance cases where we canim- There are many -- HamsaBalakrishnan Hamsa Balakrishnan migrated from the nuts-and-boltsofhow aircraft as anaerospace engineer. Over time,her research Technology (MIT).Balakrishnan began hercareer and astronautics at the Massachusetts Institute of Hamsa Balakrishnanisaprofessor ofaeronautics Image credit:HamsaBalakrishnan,MIT shown to helppredict flight delays intheairtraffic network. Characteristic delay states oftheNational Airspace System, identified through clustering. These delay states havebeen Air Traffic Control - H optimize pushback,orthe rate at which instructions used to solve aproblem) to on algorithms (aset ofstep-by-step range ofconditions. Themodels focused people whorunthemoperate undera realistic modelsofhow airportsandthe (MIT) analyzed historical data to develop Massachusetts Institute of Technology A team ledby HamsaBalakrishnanat the models ofairtraffic control systems. new embeddedtechnologies andbetter on therunway through acombination of the wait timethat planes spend taxiing Researchers are working to helpreduce excess fuelandcosts money. only isitinconvenient, but italsoburns seems like forever before taking off?Not Algorithms toReduceAirTraffic Congestion away from thegate andsitsfor what ave you beenonaplanethat pulls Who does thisstuff NAS delaystate1 NAS delaystate4 NAS delaystate3 NAS delaystate4 runways moving efficiently. tools andalgorithms required to keep flights safe and management and works to comeup with the analytic ate overall. Today, shestudies air traffic control and fly to thedetails ofhow airtraffic systems oper 100 100 NAS delaystate2 NAS delaystate5 NAS delaystate1 NAS delaystate6 Uber, thepopularridesharing service. and helpimprove thequeuingsystem of used to study energy usage inahome of research. Related algorithms are being beyond airtraffic control toother areas these modelsandalgorithms extends The fundamental research that underlies delays to thesystem. reduce taxiing time,withoutaddingany This prototyping showed that they could ers to test underreal airportconditions. Administration (FAA) andmajorair carri took theirmodelsto theFederal Aviation the algorithms. Then,theresearchers periments, to ensure theirvalidity of team ran simulations, orvirtual ex After themodels were developed, the planes shouldleave theirgates. 100 100 NAS delaystate3 NAS delaystate6 NAS delaystate5 NAS delaystate2 ? - 100 100 - -

NOV 10, 2015 • VOLUME 04 • ISSUE 02 An algorithm is a set of ste-by-step instructions used to solve a problem. As with many problems, there is not only one solution, and you must make decisions about tradeoffs such as time, simplicitiy, ACTIVITY or cost. The following activity, adapted from Curriki.org, will help to provide an introduction to the concept of algorithms.

1. Make a list of the different transportations options that you have to reach home from your current loca- tion. You might walk, take a bus or taxi, or drive. 2. Develop a simple algothim for each (e.g., Step 1. Go to the bus stop. Step 2. Board the bus. Step 3. Get off the bus at the closest stop near your home. Step 4. Walk to your home.). 3. If you were running late to meet someone at your home, which option/algorithm would be optimal? If you suddenly realized that you forgot your wallet, which option would you select and why? 4. As you can see, under different circumstances, one option/algothm may be optimal over another. In Dr. Balakrishnan’s work described here, what was the factor used to optimize the algorithm she and her team developed?

Learn More About CS Bits & Bytes is a bi-weekly newsletter Designing tomorrow’s air traffic control systems highlighting innovative computer science research. It is our hope that you will use http://go.usa.gov/cr3Qm CS Bits & Bytes to engage in the multi- faceted world of computer science to become not just a use, but also a creator MIT’s Hamsa Balakrishnan of technolgoy. Please visit our website at: www.nsf.gov/cise/csbytes. http://www.mit.edu/~hamsa/research.html

Aircraft queues on the surface of New York’s LaGuardia airport, as seen from surface surveillance data. The green icons depict departures, and the red ones denote arrivals. Note the long queue of departures waiting to takeoff from the runway. Visualization using Google EarthTM . Image credit: Hamsa Balakrishnan, MIT

National Science Foundation CS Bits & Bytes Figure 4-7: Queue for conﬁguration 31 4 http://www.nsf.gov/cise/csbytes Computer & Information Science & En| gineering Directorate 4201 Wilson Blvd, Suite 1105 Please direct all inquiries to: 4.4.2 Taxi TimesArlington, VA 22230 [email protected]

The average departure taxi times for configuration 31 4 are shown in Figure 4-8. This | configuration also has a peak in the morning around 0900 hours, with an average taxi time of 27 minutes. However, this configuration does not have the same increase in values over the afternoon and evening periods. The average taxi time across all departures in this configuration is 21.9 minutes. The arrival taxi times are shown in Figure 4-9. The average taxi time for all arrivals under this configuration was 6.1 minutes.