Modelling and Simulating Airport Surface Operations with Gate Conflicts Shannon Zelinski Robert Windhorst NASA Ames Research Center Royal Aeronautical Society and AIAA Flight Simulations Conference, London 13-14 November 2017
SOSS is: Surface Operations Scheduler & Simulator (SOSS) A fast-time simulation environment for surface operations Used to develop and test surface scheduling concepts Currently testing a surface scheduling concept for Charlotte Douglas International (CLT) 2
CLT Surface Operations Challenges Complex runway constraints 3
CLT Surface Operations Challenges Complex runway constraints Limited space for taxiing 4
CLT Surface Operations Challenges Complex runway constraints Limited space for taxiing Heavy use of limited gates 5
Gate Conflicts Flights need the same gate at the same time: Arrival is early Departure is late or held for metering Common in hub operations arrival/departure banks demand dep arr time Resolution option: Temporary parking in hardstands 6
Objectives Describe SOSS and new functionality to model hardstand operations Compare gate conflict management approaches impact on surface scheduling operations 7
Outline SOSS Gate Conflict Management Experiment Setup Results 8
SOSS Airport Model Runways Active Movement Area (AMA) Ramp Gates 9
SOSS Airport Model Dep Queue Node Departure Node Crossing Node Arrival Node Spot Node
SOSS Airport Model Dep Queue Node Departure Node Hardstand Nodes Crossing Node Arrival Node Spot Node 11
Flight Taxi Movement and Routing 12
Scheduler Interface Flight states and intent Reroutes Scheduler Release times Gate nodes Hardstand nodes 13
Outline SOSS Gate Conflict Management Prediction Resolution options Management approaches Experiment Setup Results 14
Gate Conflict Prediction target takeoff time Departure Node Gate Node pushback ready time departure target gate release time time 15
Gate Conflict Prediction Arrival Node arrival landing time target takeoff time Departure Node Gate Node pushback ready time departure gate IN time target gate release time gate time separation time 16
Arrival Node arrival Gate Conflict Prediction Predicted Gate Conflict landing time Departure Node Earliest arrival Target departure gate release gate IN < + b target takeoff time Gate Node pushback ready time departure gate IN time target gate release time b time 17
Arrival Node arrival Gate Conflict Resolution Resolution Options landing time Departure Node Departure Early Release Departure To Hardstand Arrival To Hardstand target takeoff time Gate Node pushback ready time departure gate IN time target gate release time b time 18
Gate Conflict Resolution Arrival Node arrival landing time Departure Node Departure Early Release target takeoff time Gate Node pushback ready time departure b gate IN time time 19
Gate Conflict Resolution Arrival Node arrival landing time Departure To Hardstand hardstand release time target takeoff time Departure Node Hardstand Node pushback ready time departure Gate Node b gate IN time time 20
Gate Conflict Prediction Arrival Node landing time Departure Node arrival Arrival To Hardstand hardstand release time target takeoff time Hardstand Node pushback ready time departure Gate Node target gate release time b gate IN time time 21
Gate Conflict Management Approaches Management Approach No Hardstand Departure Hardstand Arrival Hardstand Dual Hardstand Departure Early Release Resolutions Allowed Departure To Hardstand Arrival To Hardstand 22
Outline SOSS Gate Conflict Management Experiment Setup Results 23
Experiment Setup arrivals SOSS 0.5 sec time step departures Ops per 15-min 30 20 10 0 Demand Scenario departures arrivals 0 30 60 90 120 150 180 210 240 Simulation time (minutes) Surface congestion uncertainty modelled Scheduler Called every 10 seconds Surface metering ON Gate conflict b = 5 min Gate conflict management (4) 24
Outline SOSS Gate Conflict Management Experiment Setup Results Resolution types Gate time separation Runway time predictability Surface transit time 25
Results: Resolution Types Dual Hardstand Arrival Hardstand Departure Hardstand No Hardstand None Arrival To Hardstand Departure Early Release Arrival and Departure To Hardstand Departure To Hardstand 0 1 2 3 4 5 6 7 8 9 10 11 12 13 Number of gate conflict flight pairs 26
Results: Gate Time Separation separation violation Dual Hardstand Arrival Hardstand Departure Hardstand No Hardstand None excess separation Arrival and Departure To Hardstand Arrival To Hardstand Departure To Hardstand -5-4 -3-2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 actual gate separation - b Departure Early Release 27
Results: Gate Time Separation separation violation Dual Hardstand Arrival Hardstand Departure Hardstand No Hardstand None excess separation Arrival and Departure To Hardstand Arrival To Hardstand Departure To Hardstand Arrival resolutions achieve more desired gate time separation -5-4 -3-2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 actual gate separation - b Departure Early Release 28
Results: Runway Time Predictability Error (min) 3 2 Runway Time Prediction Error at Ready Time Average Standard Deviation late early 1 0-1 -2-3 No Departures involved in gate conflict Other departures Dep Arr Dual No Dep Arr Dual 29
Results: Runway Time Predictability Error (min) 3 2 Runway Time Prediction Error at Ready Time Average Standard Deviation late early 1 0-1 -2-3 No Arrival resolutions have least impact Departures involved runway in gate time conflict predictability (13) Other departures (186) Dep Arr Dual No Dep Arr Dual 30
Results: Surface Transit Time Average transit time (min) Departures (time between ready and takeoff) 20 18 16 14 12 10 8 6 4 2 0 Flights involved in gate conflict Other flights No Dep Arr Dual No Arrivals Little difference in surface transit times for others Dep Arr Dual 31
Results: Surface Transit Time Average transit time (min) Departures (time between ready and takeoff) 20 18 16 14 12 10 8 6 4 2 0 No Dep Arr Arrival resolution greatly impact arrival transit times Flights involved in gate conflict Other flights Dual Arrivals (time between landing and gate) No Dep Arr Dual 32
Summary and Conclusions Gate Time Separation Runway Time Predictability Surface Transit Time Arrival resolutions are best at achieving desired gate time separation Arrival resolutions have least impact on runway time predictability Arrival resolutions greatly impact arrival surface transit times Arrival Hardstand approach is sufficient for simulations of tactical surface metering Dual Hardstand approach may be needed for simulations with large departure delays due to Traffic Management Initiatives 33
Future Work Explore use of Dual Hardstand approach in simulations with Traffic Management Initiatives Enhance SOSS to allow flights to be rerouted at any time 34
Questions Shannon.j.zelinski@nasa.gov 35