PLANET Technology
Information Day
Prague, May 26, 2003
|
|
14:30
- 15:20 |
Henk Hesselink
(National Aerospace Laboratory, The Netherlands)
AI Planning
in Practice at Airports: Departure Management
- Abstract
- Current day airport
traffic is constrained by many rules and regulations. It is the task of
the airport traffic controller to optimise the traffic flow, whilst ensuring
safety. Departure management is the complex task of optimising departure
traffic at airports involving many actors, such as controllers, pilots,
handlers, push-back, catering, fueling, and of course passengers. A departure
management function supports all actors and takes preferences of all actors
involved into account.
We used constraint reasoning for solving the departure scheduling problem.
Where aircraft are represented by the flight object, the airports regulations
are represented in constraints. The most important constraints concern speed
and time separation. Aircraft following each other are constrained by their
relative speeds and by the wake vortex, which requires a larger separation
for a small aircraft following a large one. Other constraints are the air
route that aircraft follow (i.e. optimal use of airspace through separating
following aircraft on different routes) and the acceptance rate of the air
traffic controller of the next control sector. Overconstraint problems are
solved by constraint relaxation.
The departure scheduling management function is a decision support aid for
the airport traffic controller. Especially during peak periods and in changing
meteorological conditions (e.g. changing a runway configuration), the function
has proven its use. Furthermore, environmental rules often impose a strong
constraint on the routes for aircraft to fly and controllers cannot always
oversee the consequences of certaint choices, especially when these contradict
the optimisation criterion. Trials have been performed at Prague, Hamburg,
Paris Orly, and Rome airports.
- Biography
- Henk Hesselink is employed
at the National Aerospace Laboratory in the Netherlands since 1991. He is
currently group leader "Airport Decision Support Systems". Henk
has been working in the field of air traffic control since his employment
at NLR and is involved in airport ground movement and planning since the
early days of the "Advanced Surface Movement Guidance and Control Systems"
(A-SMGCS) in 1996. He was project leader for several projects on airport
traffic controller support in planning air traffic, mainly for planning
of departing traffic from the gate, over the airport, to the first air sector.
He was also project leader for constructing the first international airport
simulator that was equiped with an A SMGCS. Currently, he is working on
the project Triple-I: Intelligence Instead of Infrastructure, a proof-of-concept
project for demonstrating new airport functions in the NLR Tower Research
Simulator.
© 2003, Web design by
Roman Barták