The Rules of the Run Cut Part III - Changing Rules and Managing the Number of Duties
Now that we have defined the duty types and created our rule version , we are in a position to run the CrewOpt program in Hastus. Because the rule version usually has rules mandated by union labor contracts, it is normally not changed except before and during contract negotiations when different proposals can be costed out. An exception might be that it could be worthwhile to try relaxing certain restrictions to beyond the minimums stipulated in the contract - always keep in mind that the more flexible you make the rules the better the solution Hastus can generate.
One of the major ways that we can reach the most cost-effective solution for the crew schedule is to manage the number of duties the schedule creates. The first time I run the CrewOpt program I input no constraints on the program. Each additional time I run the CrewOpt program I change the total number of duties, either by adding or subtracting one from the previous total, to see what effect such a change has on the overall hour cost of the solution. Of course, sometimes there are factors beyond mere cost that have to be taken into account during this process. For example, if the agency has relatively few drivers ( perhaps the agency needs to relax their requirements ) then I will be looking for a run cut that has fewer duties; conversely, if my agency is flush with operators then I will be looking for a run cut that has more duties. I can also modify the number of straight versus split duties spit out by the run cutter, although in general for a given number of duties Hastus will create as many straight runs as it is cost effective to do so.
Note that in some cases it may not be possible for Hastus to create a workable run cut that fulfills every demand you place upon it. In such cases dreaded free duties may result. Due to the complex nature of the operation, it is generally not cost-effective to attempt to manually adjust the resulting duties so that all of them become valid, but feel free to spend time trying!
In most cases you will want to refrain from changing the vehicle schedule once you have moved along in the process to the run cut. However, if a particular block cannot be cut then you may have no choice but to do so. In addition, if you allow Hastus to use pull reliefs and to "remake blocks", then the program, if necessary in order to create a cheaper solution, will change the existing blocking to have reliefs consisting of drivers pulling out buses.
In the end, you will likely have a lot of different run cuts varying from each other by as little as a couple of hours. Lest you think such small differences can have no effect, consider how much a transit agency would save if they saved one hour on weekdays for a six month bid, which consists of about 125 weekdays. Total savings for that service change, if service cost $100 per hour to provide, would be equal to 125 * $100 = $125,000 per hour. While perhaps this savings may not be enough to avoid service cuts or fare increases, it is more significant than what might have been apparent at first glance.
Once you have selected your preferred run cut selection for each day type - at least weekday, Saturday and Sunday - you should carefully go over each run to make sure they are good ones for coach operators. In particular, I always look to make sure there is enough time for operators working split runs to get between their two pieces.
While most transit agencies only work on run cuts during the applicable time in the service change process, the most effective crew schedules can often come about during the "off season" when schedulers can experiment with tweaking rule versions to increase efficiency. When every hour saved in this process can generate annual savings of $100,000 + , it is always a good time investment to spend time refining the process.