It is all about having control of your resources. How much can you deliver with the crew and the fleet that you have available? And when are too many commitments starting to jeopardise the operation?

Forecasting
It is never attractive to cancel a flight. Especially not when the passengers are waiting at the gate ready to board. Still this happens.

In retrospect it might have been better to decline the marketing department's request for additional capacity last month. But refusing profitable flights requires indisputable proof of high costs or obvious stability problems.

By letting the optimizer create a scenario with the additional flights, the implications can be accurately forecasted and the right decision can be made.

Finding the right scenario
The ability to produce a large number of scenarios makes it possible to focus on general plan properties, such as cost, stability, quality and service level, instead of on specific flight combinations. Relevant key figures are easily retrieved by creating scenarios with minimal cost, maximum crew quality, maximum stability, etc.

Creating maximal and minimal scenarios makes it possible to calculate exactly how much more production that is theoretically possible to add. It is also possible to calculate the optimal trade-off points between properties like punctuality and service level or between crew quality and cost.

Crew influence
Allowing crew to influence their rosters is an important issue for the crew members. At the same time it can cause some major problems in the planning process in terms of higher cost and lower stability. To control this trade-off it is possible to specify exactly what level of impact crew influence concepts such as bidding, fairness and bid-lines are allowed to have in terms of cost, stability and service level.

Without speed you end up with the perfect solution to the wrong problem
Access to optimization and the ongoing development of the technology have made it possible to create plans more quickly. Therefore you can start planning later with more up-to-date information. Late schedule changes can also easily be merged into the ongoing planning process.

Using the modelling language Rave you can also reduce the lead-time of process changes. It becomes easy to evaluate and introduce process changes, such as publishing time off patterns instead of exact flight information, increasing the coordination between crew planning and critical aircraft connections or allowing passenger forecasts to directly influence crew planning.

The potential for a smooth change process is enormous, and makes it possible to adjust the planning process to change with reality instead of trying to prevent the reality from changing.