(Continued from the previous issue of the FRM News Flash)
Lars, could you please explain the overall process for providing the crew with influence over their roster? When and how are requests and preferences dealt with in the crewing process?
- Certainly. Requests, as explained earlier, are guaranteed to the crew in advance and are, therefore, straightforward to manage from a planning perspective. They are allocated to the rosters before the optimization phase. The challenge with requests lies in preserving them throughout the process during roster maintenance and on the day of operation. Another challenge is to ensure they don’t adversely affect the ability to cover production or other crucial objectives, a concern operators may face if too much is conceded early in the process. For this reason, there are rules on how requests can be placed relative to other objects on the roster, validated upfront when the crew places them. Preferences are more intricate. To start with, there are two different philosophies, Fairshare and Strict Seniority. Let’s focus on the Fairshare philosophy here. A common approach is then to let the crew express priority between their different preferences by allocating ‘bid points’ to them. And…
…let me interrupt you there, Lars. Are these points limited in supply? If not, could it not be the case crew just throws in enormously high points on something, outcompeting the wishes of others?
- I was coming to that… The points are only there to express the priority between the preferences for the crew member him/herself. So, placing 100 points on having the 14th off and 200 points for a layover in London on the 14th would be the same as placing 10 and 20 points. It’s the relative difference that is important.
Hang on… surely you cannot ask for a day off AND a layover at the same time?
- Yes, you can. An example may be that either you prefer working your way to London, perhaps to see a football game, or you will have time off allowing you to go there yourself on your spare time. Conflicting requests are often okay. An operator may have very few pairings with London layovers but could easily arrange for many crew to have that day off.
I see. How is that dealt with in the optimization process then?
- The usual approach for our customers is to do a first pre-run creating ‘reference rosters’ for all crew. During this run, the optimizer is allowed to overbook the pairings, meaning that 10 captains may get that London layover. In our original example above, that layover would now add 200 points to that crew member’s total fulfilled ‘reference points’. Later, in the real production run that will be published, the pairings are, of course, not overbooked, and only one captain will get that layover. During the production run, the optimizer strives to maximize the ratio between actual points in the production run and the points given in the reference run. In a sense, the reference run represents the maximum points that the crew could reasonably achieve if the crew member was not competing with other crew for the activities assigned by the optimizer.
So, it could be seen as a ‘dream roster’?
- Not really. We refer to it as a ‘reference roster.’ It needs to contain a reasonable amount of duty days, block hours, etc. If that were not the case, the crew could ask for 30 days off in a month, and the score from that would be seen as the reference. It is up to each operator to decide what should be a reasonable amount.
OK… I was not aware that it was this advanced. What about seniority? Does it not have any impact?
- That would also be up to each operator to decide. There may often be a minimum target for the bid award ratio that differs with seniority so that, for example, the junior crew has a target to exceed 60%, and the most senior 80%. This is configurable and differs a lot in how it is dealt with. Some operators have seniority reflected in the requests only, and treat preferences fairly ‘on top’ of those. Whereas others use preferences only and do or don't have seniority influencing the bid award ratio target.
And what if the target cannot be reached?
- Should that happen, perhaps due to crew competing over the same pairings, it will normally be adjusted for in the next planning period. A crew member reaching 70%, when the target was to exceed 80%, will have the target elevated next month to make up for the difference. Just so that the same individual does not, by chance, become unlucky month after month.
Thank you, Lars. Any final reflections on this or the importance for FRM?
- I would think that the impact on FRM is huge. People differ in physiology and also have different family situations, commute times, hobbies, etc. Allowing the crew to influence their rosters at scale is, in my opinion, crucial for long-term satisfaction and well-being. Almost all of our rostering customers provide this capability to the crew, and although, due to the preconditions, not all wishes can be granted, we do know it delivers high overall influence to the crew in a balanced way so that other objectives, such as staffing all flights, still can be secured. Also, if you think it sounds somewhat complex to deal with preferences, as I explained it above, I also like to add that we are providing what we call ‘Lifestyle Bidding’. This bidding model focuses on roster structures rather than trip properties, so that a crew member can ask for a ‘commuter roster’ for example, or a ‘nights-at-home roster’. More about that another time perhaps? And maybe also talk about strict-seniority rostering, which works quite differently? :-)
Indeed another time, Lars. Thank you again!
Please find more information about Jeppesen Crew Rostering here.
Comments