Scheduler

Input

• surface data (airport state)
• request (start location, end location, scheduled time, aircraft)

Model Constraints

• separation
• temporal (start times/ duration)
• aircraft
• connectivity (two locations connected by a path)
• airport configuration (runways can be used for either arrival or departure and can be switched)

Output

Output of planner:

• routes: [(e1,t1) (e2,t2) ...]

Output of scheduler:

• spot release (hold/go)
• push back time (hold/go)

Performance

• minimize queue size at runway
• minimize taxi time
• or, others

note

• taxi time: from pushback to take off
• it's better to stay at gate if delayed when making push back decision
• run way could be use in some specific direction and it may change
• dynamic scheduling is more realistic
• how often scheduling is happened, and what is the window size