National Academies Press: OpenBook
« Previous: Chapter 5 - Runcutting
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 299
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 300
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 301
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 302
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 303
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 304
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 305
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 306
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 307
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 308
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 309
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 310
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 311
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 312
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 313
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 314
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 315
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 316
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 317
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 318
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 319
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 320
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 321
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 322
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 323
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 324
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 325
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 326
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 327
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 328
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 329
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 330
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 331
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 332
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 333
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 334
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 335
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 336
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 337
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 338
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 339
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 340
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 341
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 342
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 343
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 344
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 345
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 346
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 347
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 348
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 349
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 350
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 351
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 352
Suggested Citation:"Chapter 6 - Rostering." National Academies of Sciences, Engineering, and Medicine. 2009. Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling. Washington, DC: The National Academies Press. doi: 10.17226/14257.
×
Page 353

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Chapter 6. Rostering Ͳ.ͭ Basic Rostering (Level ͭ) Ͳ.ͮ Intermediate Rostering (Level ͮ) Ͳ.ͯ Advanced Rostering (Level ͯ)

Chapter 6. Rostering 6-3 6.1 Basic Rostering—Rosters and Their Purpose Rostering is the process of grouping daily operator runs into packages of weekly work assignments. Operators are generally given the opportunity, based on order of seniority, to pick their work for the next period of time. This period of time is often called a sign-up period or bid period. At most agencies, sign-ups generally occur three or four times a year, so the bid period might last for three or four months. Rosters have a list of work assignments, indicated in the example below by the roster number. Assignments may include mixtures of runs, such as: Weekday runs only Weekday runs and a Saturday run Weekday runs and a Sunday run Weekday runs, a Saturday run, and a Sunday run Roster No. Sun Mon Tue Wed Thu Fri Sat WeeklyPay Hours 1001 Off 1018:00 1018:00 1018:00 1018:00 1018:00 Off 40:00 1002 Off 1028:00 1028:00 1028:00 1028:00 1028:00 Off 40:00 1003 70110:00 10310:00 10310:00 10310:00 Off Off Off 40:00 1004 Off Off 10410:00 10410:00 10410:00 10410:00 Off 40:00 1005 Off 1058:00 1058:00 1058:00 1058:00 1058:00 Off 40:00 1006 Off 10410:00 Off Off 10310:00 10310:00 60110:00 40:00 In the example, roster numbers are four digits and all run numbers are three digits. Weekday runs begin with ͭ, Saturday runs begin with Ͳ, and Sunday runs begin with ͳ. Each agency has its own numbering convention. Weekly work assignments typically include fi ve daily runs, each close to or over eight hours. Some agencies use ͭͬ-hour daily runs, in which case the roster would also include assignments with four daily ͭͬ-hour runs (see Roster No. ͭͬͬͯ and ͭͬͬͰ in the example). Normally, these runs are not allowed to go too far over ͭͬ hours per day. • • • • LEVEL 1 rostering The process of grouping daily op- erator runs into packages of weekly work assignments. The fi nished product is known as a roster or a bid package. sign-up The process in which operators se- lect work assignments. Most agen- cies have three or four sign-ups each year. Sign-up is also called “bid,” “line-up,” “pick,” “shake-up,” and “mark-up.” sign-up or bid period The period of time that a specifi c sign-up is in effect, usually three or four months. Level ͭ. Basic Rostering

Level ͭ. Basic RosteringChapter 6. Rostering 6-4 Rosters generally remain in eff ect throughout the sign-up period. In the event that a roster is permanently vacated, the agency may hold a line pick in between scheduled sign-ups to fi ll the roster. Typically, only operators with less seniority than the prior holder of the vacated roster are eligible to bid. A line pick may also be held on one route only if there have been signifi cant schedule changes to the route during the sign-up period. Minor changes to a trip or series of trips made during the sign-up period (due to detours or minor running time problems) are usually addressed by means of a “patch” or temporary modifi cation. An agency’s labor agree- ment may restrict the conditions under which a patch can be used. Larger agencies operating out of multiple garages may also hold a system sign-up. This pro- vides operators an opportunity to transfer to another garage. Typically, one scheduled sign-up during the year may be designated as a system sign-up. Types of Rosters Building a roster of weekly work is carried out in one of two ways, depending on agency policy and work rules: agency-developed rosters, built by the schedulers operator-developed rosters (also known as cafeteria-style rostering), assembled by individual operators at the time of bidding based on their seniority Agency-Developed Rostering The roster shown above is an example of an agency-developed roster. The agency packages daily runs and days off into a weekly work assignment. Operators then pick their work from the list of weekly rosters. When developing a roster under this scenario the scheduler is trying to achieve a number of outcomes, including: Meet the requirements of the labor agreement. As always this is the highest and most stringent consideration. Typically the labor agreement will mandate areas including number of days worked, minimum/maximum weekly (or period) hours, days off require- ments, minimum rest breaks, run type limitations, etc. Achieve high levels of effi ciency. This outcome can be defi ned in a number of ways but probably will include minimizing operator requirements (defi ned as roster lines), and minimizing pay hours. Meet operator preferences inclusive of days off patterns, types of runs operated, rest breaks, and other working conditions. • • • • • line pick A sign-up held in between sched- uled sign-ups to fi ll one or more runs permanently vacated due to illness, disability, or termination. Only operators with less seniority than the prior holder of the vacated run are eligible to bid. Also, a special sign-up held on one route only due to a signifi cant schedule change during the sign-up period on the route. patch A temporary modifi cation to a trip or series of trips on a route imple- mented during the sign-up period to account for a detour or to address minor running time problems. system sign-up A scheduled sign-up during which operators may transfer from one garage to another. System sign- ups are usually held no more than once a year. At intermodal agen- cies, the system sign-up may allow an operator to transfer between modes as well.

Chapter 6. Rostering 6-5 Level ͭ. Basic Rostering As is often the case, the scheduler is trying to balance effi ciency requirements against pre- ferred working conditions. Agency-developed rosters may be simple one-run patterns or may be complex combinations of runs, days off , and patterns. Cafeteria-style or Operator-Developed Rostering In cafeteria-style rostering, an operator can choose both specifi c daily runs and days off from a master list of all runs and all available days off . The name comes from how customers in a cafeteria choose individual food items that make up their meal. The schedule department will develop and post a Master Run List containing all week- day, Saturday, and Sunday runs and a master list of available days off for the sign-up. In re- ality the schedule department is primarily an administrator of the rostering process under this scenario, as opposed to having responsibility for cre- ation or development of the rosters. The example of master lists shown below uses the same numbering scheme for runs as in the previous example. The list of available days off shows the total number of days off that may be chosen for each day of the week. Note that in the master lists, runs and days off are listed separately for four-day and fi ve-day rosters. Some transit systems that use cafeteria ros- agency-developed rostering The process in which the transit agency packages daily runs into weekly work schedules or rosters in advance of the sign-up. The opera- tors then select from the prepared rosters. cafeteria-style rostering The process in which operators create their own rosters by select- ing daily runs and days off from a master list. master run list A list containing all weekday, Saturday, and Sunday runs. The Master Run List may be the Run Guide or some variation of the Run Guide. Typically used in cafeteria rostering. posting Notifi cation to operators of all work assignments that will be available for selection during the next sign- up. Runs are posted for cafeteria rostering; rosters are posted for agency-developed rostering. Runs and rosters are posted for a number of days prior to the start of actual bidding to provide time for opera- tors to study their options prior to making their selection. MASTER RUN LIST AND MASTER DAYS OFF LIST Weekday Runs Saturday Runs Sunday Runs Run # Pay hours Run # Pay hours Run # Pay hours 101 8:00 601 10:00 701 10:00 102 8:00 602 8:45 702 8:00 103 10:00 603 8:15 703 8:21 104 10:00 604 8:00 105 8:00 106 8:22 107 8:45 Available Runs - Five Day Rosters Sun Mon Tue Wed Thu Fri Sat 702 101 101 101 101 101 602 703 102 102 102 102 102 603 105 105 105 105 105 604 106 106 106 106 106 107 107 107 107 107 Available Days Off - Five Day Rosters Sun Mon Tue Wed Thu Fri Sat 4 1 1 1 1 1 3 Available Runs - Four Day Rosters Sun Mon Tue Wed Thu Fri Sat 701 103 103 103 103 103 601 104 104 104 104 104 Available Days Off - Four Day Rosters Sun Mon Tue Wed Thu Fri Sat 2 1 1 1 1 1 2

Level ͭ. Basic RosteringChapter 6. Rostering 6-6 tering and have both ͱ/ʹ and Ͱ/ͭͬ rosters have agreements with their unions to use agency-de- veloped rosters for the Ͱ/ͭͬs, because cafeteria rostering does not always work well for Ͱ/ͭͬs. In addition, a master list of available extraboard (relief, stand-by) assignments is posted along with a master list of eligible operators. The operator list is typically sorted in order of descend- ing seniority and includes the day and time of each operator’s turn to pick listed next to his or her name. Picking runs and days off can be as simple as initialing the desired run and days off on the master lists. Schedule and/or operations department personnel, and sometimes Union personnel, usually “offi ciate” during the sign-up to ensure that operators pick work according to the seniority list and that all rules governing the cafeteria roster and sign-up process are followed. There are several variations within cafeteria-style rostering, often based on specifi c constrain- ing elements of the labor agreement or past practice. Constraints can include: Days off must be consecutive unless consecutive days are no longer available. Routes cannot be mixed during the work week if the same route is available for all days. Run types (straight/split) cannot be mixed until necessary to form a full weekly assignment. A minimum number of off -duty hours must separate each run picked. Implications of Each Approach Agencies that use cafeteria rostering believe that operators are more conscientious about the quality of their work when they have more control over their work hours and assignments. This can translate into non-quantifi able benefi ts such as good customer relations as well as into quantifi able benefi ts such as lower absenteeism, reduced worker compensation claims, and fewer accidents. Under cafeteria rostering, senior operators often pick runs greater than eight hours in length to qualify for more overtime pay. Since senior operators usually earn the most and pick fi rst; this can increase agency costs and leave junior operators with runs that pay eight hours or less. As a result, some senior operators earn high levels of weekly overtime while junior operators are left with weekly assignments that pay less than Ͱͬ hours. If there is a Ͱͬ-hour paid weekly guarantee, the transit agency ends up paying more make-up time. A signifi cant issue related to cafeteria-style rostering can be that while the senior operators (i.e., those who get to pick the work fi rst) are able to achieve their preferences, the less senior operators tend to have the “leftover” less-desirable work days from which to construct their • • • • extraboard A group of operators who provide coverage of vacant runs and other work on a daily or weekly basis. Operators may pick the extraboard during a sign-up or may be as- signed to the extraboard if no more runs are available. stand-by time Time that an operator spends at the garage at the agency’s direction awaiting assignment of a run or a piece of work. Usually associated with a report by an extraboard operator, stand-by is intended to provide a pool of operators that will be available to fi ll runs vacated by unscheduled absences. make-up time Time added to an operator’s work hours to bring the total up to the guaranteed minimum (usually eight hours per day or 40 hours per week). Full-time operators often have an eight-hour guarantee, even if their runs are short of eight hours. If your property uses a cafeteria-style bid, it is important to monitor to ensure that the runs being selected are “legal” and follow the rules established by your contract and practice. Errors allowed in bidding may result in costly rebids. Tip

Chapter 6. Rostering 6-7 Level ͭ. Basic Rostering workweeks—late runs, weekend work, lower-paying runs, etc. This can lead to a trend of higher turnover of less senior staff . Systems that use agency-developed rostering argue that preassembled rosters can be devel- oped in a more cost-eff ective manner and thus save money. This is especially true for agencies with no daily guarantee of eight hours pay and with weekly instead of daily overtime pay. In these cases, the agency can combine runs of over eight hours with runs of under eight hours into a weekly assignment of Ͱͬ hours, thus minimizing overtime and make-up pay. The example below shows potential cost savings for an agency using agency-developed roster- ing without a daily guarantee and daily overtime. One long run of ʹ:Ͱͬ can be combined with four smaller runs to achieve a weekly assignment of Ͱͬ hours with no overtime and no make- up pay. Savings can be accrued even if overtime is calculated daily. In the example, if daily overtime is paid (at time and a half) but there is no daily guarantee, then total pay hours would be Ͱͬ:ͮͬ. ROSTER WITH DAILY MAKE-UP AND DAILY OVERTIME Time Sun Mon Tue Wed Thu Fri Sat Weekly Total Platform Off Off 7:50 7:50 7:50 7:50 8:40 Guarantee 0:10 0:10 0:10 0:10 Overtime 0:20 Pay Hours 8:00 8:00 8:00 8:00 9:00 41:00 ROSTER WITH WEEKLY MAKE-UP AND WEEKLY OVERTIME Time Sun Mon Tue Wed Thu Fri Sat Weekly Total Platform Off Off 7:50 7:50 7:50 7:50 8:40 Guarantee Overtime Pay Hours 7:50 7:50 7:50 7:50 8:40 40:00 In the absence of daily guarantees and overtime, savings are possible with cafeteria rostering. However, these savings would occur randomly and would not reach the potential savings as- sociated with agency-developed rostering. An agency can achieve signifi cant cost savings using agency-developed rostering if make-up pay and overtime pay are calculated on a weekly basis. Tip Many transit agen- cies roster the work for drivers in advance and let drivers “bid” or pick work from runs that have already been packaged into a weekly work schedule or roster. Other agencies let drivers pick in- dividual daily assignments. If your agency rosters work in advance, it is important for you to know the policy on the number of straight and split runs, consecutive days off, and other work rules that impact the roster. Tip

Level ͭ. Basic RosteringChapter 6. Rostering 6-8 Roster Examples This section provides examples of rosters using the Route ͵ͳ example from previous chapters. Both agency-developed and cafeteria rosters are shown. The basic section of the runcutting chapter (Chapter ͱ) prepared a run guide for Route ͵ͳ. This will serve as our starting point in developing rosters. The run guide is shown below. Option 1 - Completed Run Cut Time Place Type Time Place Type 101 Str 2 5:46 6:01 Garage Pull 12:09 A Street 12:34 6:08 0:30 0:10 6:48 6:48 1:12 0:00:00 0:00:00 8:00:00 1.304 102 Str 2 11:44 12:09 A Street 18:38 Garage Pull 18:53 6:29 0:30 0:10 7:09 7:09 0:51 0:00:00 0:00:00 8:00:00 1.234 103 Str 3 5:35 5:50 Garage Pull 12:39 A Street 13:04 6:49 0:30 0:10 7:29 7:29 0:31 0:00:00 0:00:00 8:00:00 1.174 104 Str 3 12:14 12:39 A Street 18:19 Garage Pull 18:34 5:40 0:30 0:10 6:20 6:20 1:40 0:00:00 0:00:00 8:00:00 1.412 105 Str 5 6:05 6:20 Garage Pull 13:09 A Street 13:34 6:49 0:30 0:10 7:29 7:29 0:31 0:00:00 0:00:00 8:00:00 1.174 106 Str 5 12:44 13:09 A Street 19:53 Garage Pull 20:08 6:44 0:30 0:10 7:24 7:24 0:36 0:00:00 0:00:00 8:00:00 1.188 1 5:31 5:46 Garage Pull 9:19 Garage Pull 9:24 6 14:46 15:01 Garage Pull 19:49 Garage Pull 20:04 108 Pt 4 5:50 6:05 Garage Pull 9:38 Garage Pull 9:53 3:33 0:30 0:00 4:03 4:03 0:00 0:00:00 0:00:00 4:03:00 1.141 109 Pt 7 15:05 15:20 Garage Pull 19:19 Garage Pull 19:34 3:59 0:30 0:00 4:29 4:29 0:00 0:00:00 0:00:00 4:29:00 1.126 54:32 4:50 1:00 60:21 5:21 0:35 2:09 68:27 1.255 0:50 0:00 2:09:00 11:55:30 1.428 Spread Penalty Paid Hours Pay/ Plat 107 Spl 8:21 9:11 14:18 0:00 0:35:30 Sign Off Time Plat Hours Total Hours Spread Guarantee Overtime Report Hours Travel Hours Run # Type Block # Report Time Start End Because Route ͵ͳ operates only on weekdays, available days off are on Saturdays and Sundays only. In this example and for weekday-only service in general, the number of available days off on Saturday and on Sunday is equal to the number of weekday runs that may be picked by operators. One of the important factors that rostering tells us is how many operators we need. The table below shows a formula for computing the total number of operators needed and the number of available days off by day. This table uses a hypothetical number of daily, Saturday, and Sun- day runs. Calculate the number of operators you will need. (Number of Weekday Runs x 5 + Number of Saturday Runs + Number of Sunday Runs) / 5 Day Work Week = Number of Operators Required Tip

Chapter 6. Rostering 6-9 Level Í­. Basic Rostering OPERATORS REQUIRED FOR 8-HOUR RUNS Weekly total Day # daily runs # days per week (daily runs * days) Weekdays (M-F) 120 5 600 Saturdays 54 1 54 Sundays 26 1 26 Weekly Total 680 Total Operators (weekly total runs / 5 days of work per operator) 136 DISTRIBUTION OF DAYS OFF Operators off each day Total # operators # daily runs (# operators - # daily runs) Weekdays (M-F) 136 120 16 Saturdays 136 54 82 Sundays 136 26 110 Weekly Total (weekdays * 5 + Saturday + Sunday) 272 OPERATORS REQUIRED FOR 10-HOUR RUNS Weekly total Day # daily runs # days per week (daily runs * days) Weekdays (M-F) 100 5 500 Saturdays 40 1 40 Sundays 40 1 40 Weekly Total 580 Total Operators (weekly total runs / 4 days of work per operator) 145 DISTRIBUTION OF DAYS OFF Operators off each day Total # operators # daily runs (# operators - # daily runs) Weekdays (M-F) 145 100 45 Saturdays 145 40 105 Sundays 145 40 105 Weekly Total (weekdays * 5 + Saturday + Sunday) 435

Level ͭ. Basic RosteringChapter 6. Rostering 6-10 Note that in the eight-hour example, it would not be possible to supply each operator with consecutive days off (which is a requirement in the contracts of most moderate- to large-sized systems). The reason is the diff erence between the number of Saturday runs and Sunday runs. In this case, the diff erence is ͮʹ (ͱͰ Saturdays minus ͮͲ Sundays). We have already calculated that ͭͲ operators will be off each weekday. (Note the ͭͲ operators off each weekday, or calcu- lated another way, the ͱͰ Saturday runs plus ͮͲ Sunday runs would equal ʹͬ pieces of work on weekends that must be accounted for in days off . This number is divided by ͱ, the number of days each relief operator will work each week, to equal the ͭͲ.) Since ͭͲ is less than ͮʹ, ͭͮ runs will have to have split days off . If the union contract does not allow this, the sole alternative, which is expensive, may be to detail ͭͮ runs on Saturday to the extraboard or allow them to be bid by operators on their days off at overtime. The use of part-time operators to cover the open Saturday runs may be another option, contingent upon the labor agreement. The lesson here is to make sure your transit system does not get out of balance in the amount of service provided on Saturday versus Sunday. Applying these formulas to Route ͵ͳ, with nine weekday runs and no weekend service: OPERATORS REQUIRED FOR 8-HOUR RUNS ON ROUTE 97 Weekly total Day # daily runs # days per week (daily runs * days) Weekdays (M-F) 9 5 45 Saturdays 0 1 0 Sundays 0 1 0 Weekly Total 45 Total Operators (weekly total runs / 5 days of work per operator) 9 DISTRIBUTION OF DAYS OFF Operators off each day Total # operators # daily runs (# operators - # daily runs) Weekdays (M-F) 9 9 0 Saturdays 9 0 9 Sundays 9 0 9 Weekly Total (weekdays * 5 + Saturday + Sunday) 18 We can check this total by multiplying the number of operators (͵) by the number of days off per week required by each (ͮ) to equal the weekly total of operator days off (ͭʹ). To calculate the number of days off required in a given week. Total days off = Number of Operators required x Number of days off per week In a traditional fi ve-day work week the number of days off per week is two. Tip

Chapter 6. Rostering 6-11 Level ͭ. Basic Rostering Agency-Developed Rosters As stated above, under the agency-developed rostering approach, operators pick their weekly work from a master list of weekly rosters, “Lines of Work.” An agency will typically consider several variations before deciding on the fi nal master list to post. Agency-developed rosters typically have an operator work the same workweek repeatedly, but operators can also cycle through more than one line of work over the course of the sign-up period. The fi xed workweeks (Lines of Work) may also allow the introduction of a Rotating Roster, where the operators cycle through the Lines of Work on a weekly (or other) rotation. A common convention for the format used for the master list of weekly rosters is shown here: Roster No. Sun Mon Tue Wed Thu Fri Sat Weekly Work Hours # # # # # x:xx x:xx x:xx x:xx x:xx 1001 Off Off xx:xx Under this numbering convention, a ͭͬͬͬ series number is used as the weekly roster number to avoid confusion with the numbers used to represent daily runs. “#” represents the daily run number under each day of the week, and “x:xx” represents the work hours associated with the run. Daily work hours can be summed across days to obtain weekly work hours. If preferred, pay hours can be substituted for work hours. As with many of the processes described throughout this manual the potential exists to use either “automated” or “manual” methods. Automated methods refer to use of computerized scheduling systems to develop rosters, based on a range of user-defi ned inputs. Manual in this context assumes the use of spreadsheets to assist in development of rosters. The level of sophistication of the spreadsheet depends upon user spreadsheet experience/competence. We will discuss further how rosters can be enhanced by at least some basic spreadsheet automa- tion. line of work A weekly work package, developed during rostering, that comprises a fi xed set of runs and days off for a set workweek. rotating (rotary) roster A roster where operators cycle through the weekly Lines of Work over the course of the sign-up period.

Level ͭ. Basic RosteringChapter 6. Rostering 6-12 Generally the best rosters have as many runs close to the eight-hour daily target as possible. However, every property is different, and it is always important to understand your property’s work rules before you move to rostering. Tip Our fi rst eff ort at developing a roster for Route ͵ͳ might look like this: ROSTER - Route 97 (variation 1) Roster No. Sun Mon Tue Wed Thu Fri Sat Weekly Hours Make-up Hours Overtime Spread Hours Weekly Pay Hours 1001 Off 101 101 101 101 101 Off 6:48 6:48 6:48 6:48 6:48 34:00:00 6:00 0:00:00 0:00:00 40:00:00 1002 Off 102 102 102 102 102 Off 7:09 7:09 7:09 7:09 7:09 35:45:00 4:15 0:00:00 0:00:00 40:00:00 1003 Off 103 103 103 103 103 Off 7:29 7:29 7:29 7:29 7:29 37:25:00 2:35 0:00:00 0:00:00 40:00:00 1004 Off 104 104 104 104 104 Off 6:20 6:20 6:20 6:20 6:20 31:40:00 8:20 0:00:00 0:00:00 40:00:00 1005 Off 105 105 105 105 105 Off 7:29 7:29 7:29 7:29 7:29 37:25:00 2:35 0:00:00 0:00:00 40:00:00 1006 Off 106 106 106 106 106 Off 7:24 7:24 7:24 7:24 7:24 37:00:00 3:00 0:00:00 0:00:00 40:00:00 1007 Off 107 107 107 107 107 Off 9:11 9:11 9:11 9:11 9:11 45:55:00 0:00 2:57:30 10:45:00 59:37:30 1008 Off 108 108 108 108 108 Off 4:03 4:03 4:03 4:03 4:03 20:15:00 0:00 0:00:00 0:00:00 20:15:00 1009 Off 109 109 109 109 109 Off 4:29 4:29 4:29 4:29 4:29 22:25:00 0:00 0:00:00 0:00:00 22:25:00 Total 301:50:00 26:45:00 2:57:30 10:45:00 342:17:30 This roster features consistent runs in each assignment. Rosters ͭͬͬͭ through ͭͬͬͳ are full- time rosters, while Rosters ͭͬͬʹ and ͭͬͬ͵ are part-time rosters. There is an extensive amount of make-up time in this roster, but because this example draws from only one route, there are fewer options for hooking trips and cutting runs. Are there any “tweaks” that we could make to this roster to enhance effi ciency? If the primary concern were high levels of overtime, then addition of Ͱ/ͭͬ work assignments would be worth exploring. To digress for a moment, overtime is not necessarily a bad thing, since it can mini- mize the number of operators. Arguments in favor of minimizing the number of operators include: Fewer operators translate to lower costs for benefi ts such as medical premiums and pension obligations. In good economic times when jobs are plentiful, it is often diffi cult to recruit a suffi cient number of operators. • •

Chapter 6. Rostering 6-13 Level ͭ. Basic Rostering Minimizing the number of operators translates directly to developing rosters with higher levels of overtime. The tradeoff is often seen in purely economic terms: the added cost of benefi ts for a new operator versus higher overtime costs. There is also an important argument against minimizing the number of operators: Longer work hours can potentially result in increased number of accidents (due to fa- tigue behind the wheel) and increased absenteeism. Another possible change is to mix and match daily runs to bring each roster as close to Ͱͬ weekly pay hours as possible. This alternative, shown as variation ͮ, assumes that overtime and make-up time are paid on a weekly, not a daily, basis. ROSTER - Route 97 (variation 2 - weekly overtime and make-up) Roster No. Sun Mon Tue Wed Thu Fri Sat Weekly Hours Make-up Hours Overtime Spread Hours Weekly Pay Hours 1001 Off 101 101 101 101 101 Off 6:48 6:48 6:48 6:48 6:48 34:00:00 6:00 0:00:00 0:00:00 40:00:00 1002 Off 102 102 102 102 102 Off 7:09 7:09 7:09 7:09 7:09 35:45:00 4:15 0:00:00 0:00:00 40:00:00 1003 Off 103 103 103 103 103 Off 7:29 7:29 7:29 7:29 7:29 37:25:00 2:35 0:00:00 0:00:00 40:00:00 1004 Off 107 107 104 104 104 Off 9:11 9:11 6:20 6:20 6:20 37:22:00 2:38 0:00:00 4:18:00 44:18:00 1005 Off 105 105 105 105 107 Off 7:29 7:29 7:29 7:29 9:11 39:07:00 0:53 0:00:00 2:09:00 42:09:00 1006 Off 106 106 106 106 106 Off 7:24 7:24 7:24 7:24 7:24 37:00:00 3:00 0:00:00 0:00:00 40:00:00 1007 Off 104 104 107 107 105 Off 6:20 6:20 9:11 9:11 7:29 38:31:00 1:29 0:00:00 4:18:00 44:18:00 1008 Off 108 108 108 108 108 Off 4:03 4:03 4:03 4:03 4:03 20:15:00 0:00 0:00:00 0:00:00 20:15:00 1009 Off 109 109 109 109 109 Off 4:29 4:29 4:29 4:29 4:29 22:25:00 0:00 0:00:00 0:00:00 22:25:00 Total 301:50:00 20:50:00 0:00:00 10:45:00 333:25:00 The shortest full-time work assignment in variation ͮ is ͯͰ:ͬͬ, compared to ͯͭ:Ͱͬ in variation ͭ, and no roster involves work in excess of Ͱͬ hours per week. Why does weekly overtime and make-up time make a diff erence? Variation ͯ shows the same roster as variation ͮ, but with daily overtime and make-up pay. Compare Roster ͭͬͬͳ in both examples. In variation ͮ, the combination of runs over and under eight hours results in no over- time, since weekly hours do not exceed Ͱͬ, and only ͭ:ͮ͵ in make-up time. In variation ͯ, the agency must pay overtime for Run ͳ on Wednesday and Thursday (total overtime of ͮ:ͮͮ hours, •

Level ͭ. Basic RosteringChapter 6. Rostering 6-14 which translates to an additional ͭ:ͭͭ pay hours), and must pay make-up time for the other days (ͭ:Ͱͬ on Monday and Tuesday and ͬ:ͯͭ on Friday for a total of ͯ:ͱͭ in make-up time). To- tals in variation ͯ are identical to variation ͭ. ROSTER - Route 97 (variation 3 - daily overti me and make-up) Roster No. Sun Mon Tue Wed Thu Fri Sat Weekly Hours Make-up Hours Overtime Spread Hours Weekly Pay Hours 1001 Off 101 101 101 101 101 Off 6:48 6:48 6:48 6:48 6:48 34:00:00 6:00 0:00:00 0:00:00 40:00:00 1002 Off 102 102 102 102 102 Off 7:09 7:09 7:09 7:09 7:09 35:45:00 4:15 0:00:00 0:00:00 40:00:00 1003 Off 103 103 103 103 103 Off 7:29 7:29 7:29 7:29 7:29 37:25:00 2:35 0:00:00 0:00:00 40:00:00 1004 Off 107 107 104 104 104 Off 9:11 9:11 6:20 6:20 6:20 37:22:00 5:00 1:11:00 4:18:00 47:51:00 1005 Off 105 105 105 105 107 Off 7:29 7:29 7:29 7:29 9:11 39:07:00 2:04 0:35:30 2:09:00 43:55:30 1006 Off 106 106 106 106 106 Off 7:24 7:24 7:24 7:24 7:24 37:00:00 3:00 0:00:00 0:00:00 40:00:00 1007 Off 104 104 107 107 105 Off 6:20 6:20 9:11 9:11 7:29 38:31:00 3:51 1:11:00 4:18:00 47:51:00 1008 Off 108 108 108 108 108 Off 4:03 4:03 4:03 4:03 4:03 20:15:00 0:00 0:00:00 0:00:00 20:15:00 1009 Off 109 109 109 109 109 Off 4:29 4:29 4:29 4:29 4:29 22:25:00 0:00 0:00:00 0:00:00 22:25:00 Total 301:50:00 26:45:00 2:57:30 10:45:00 342:17:30 Pay-to-Platform Rati o 1.255 Many agencies believe that it is important for an operator to work the same weekday run every day. The operator becomes more familiar with the route and also gets to know regular pas- sengers who ride at the same time every day. Agencies supporting this view are willing to ac- cept some additional cost in order to achieve route and passenger familiarity. Variation ͮ may or may not be acceptable to such agencies, because only four of the seven full-time rosters are consistent throughout the week. At other agencies, operators may prefer the variety of working diff erent runs or routes to relieve boredom (and the potential of related fatigue). The scheduler should be aware of preferences at his or her specifi c system. Union-management relations might also be a factor. The operators’ union might prefer rosters featuring work assignments with overtime, or might prefer less overtime and more work as- signments. The optimal situation is a union-preferred roster that is the least expensive among various alternatives, but agencies may be willing to accept some additional cost (with the em- phasis on “some”) to accommodate union preferences.

Chapter 6. Rostering 6-15 Level ͭ. Basic Rostering This brings us to an important point which has been emphasized throughout this manual: the scheduler should be focused on providing the best quality operational outcome for the given cost. This applies strongly to rostering where a number of trade-off s can be made (such as days-off patterns or rest times) to reduce costs, or even to provide a better quality solution within the same costs. Evaluating the Agency-Developed Rosters Choosing among roster variations can involve consideration of a number of factors. Before making a fi nal decision, an agency needs to understand fully the cost and operational implica- tions of diff erent rosters. Comparison charts are frequently used to present quantifi able information to aid in the evalu- ation of variations. An example comparison chart for the three variations of agency-developed rosters for Route ͵ͳ is shown below. COMPARISON CHART FOR THREE ROSTER VARIATIONS FOR ROUTE 97 Variation 1 Variation 2 Variation 3 Total hours for all weekly rosters 301:50:00 301:50:00 301:50:00 @ Straight time 295:55:00 301:50:00 295:55:00 @ Overtime (time and a half) 5:55:00 0:00:00 5:55:00 @ Make-up 26:45:00 20:50:00 26:45:00 @ Spread time (half-time over 10 hours) 21:30:00 21:30:00 21:30:00 Pay hours 342:17:30 333:25:00 342:17:30 Number of operators required 9 (7/2) 9 (7/2) 9 (7/2) Total platform hours 272:40:00 272:40:00 272:40:00 Pay-to-platform ratio 1.255 1.223 1.255 Variation ͭ and Variation ͯ are identical, as noted earlier. Variation ͮ is the least expensive roster, because it has no overtime and less make-up time. Variation ͮ results in a savings of ʹ:ͱͮ:ͯͬ per week. Variation ͮ also has the lowest pay-to-platform ratio. The number of op- erators is the same under all three scenarios. Computerized scheduling systems allow a more qualitative comparison of roster outcomes. Preferred parameters can be quantifi ed to check one solution against the next, without the scheduler having to wade through large tables or columns. The result can be rosters of im- proved quality which translates to an enhanced work environment for operators. pay-to-platform ratio The ratio of pay hours to platform time. For example, if an operator receives 9:00 in pay for 8:00 of platform time, the pay-to-platform ratio is 1.125 (9:00/8:00). The pay- to-platform ratio is one of the most widely used methods of measur- ing runcut effi ciency and is often used to measure the impacts of non-platform items (such as report allowance or relief allowances) on operator pay hours. Some systems use the inverse, the ratio of platform to pay hours.

Level ͭ. Basic RosteringChapter 6. Rostering 6-16 Cafeteria Rosters Under the cafeteria rostering approach, operators build their own rosters by picking their weekly work from the master run list and the master days off list. These master lists are either the actual Run Guide shown earlier or derived directly from the Run Guide. The Route ͵ͳ Run Guide was shown in the agency-developed rosters example. The calculations for required number of operators and number of available days off are identi- cal for cafeteria rosters. Runs and days off are posted, and operators make their own weekly assignment by selecting from available runs and available days off . Cafeteria rostering is simpler for the agency, since it does not have to prepare and evaluate ros- ter alternatives. However, those responsible for the sign-up process must check the numbers of available runs at the end of each day of the sign-up in order to be certain that no run was picked twice or that an odd number of work or days off were left open. The operations depart- ment usually manages the sign-up process for cafeteria rostering, so for schedulers, cafeteria rostering often means much less work (the schedules department is invariably responsible for agency-developed rosters). Agencies that use cafeteria rostering consider it a positive in terms of employee morale, since operators design their own work week. The downsides are that (ͭ) the agency cannot count on generating cost savings through eff ective rostering and (ͮ) less senior operators are often left with very undesirable work weeks. While there are many arguments pro and con, agencies with weekly but not daily guarantees, should avoid cafeteria rostering if at all possible because of the ineffi ciencies previously dis- cussed. If there are daily guarantees, then, as indicated above, the choice can be either. Agencies with weekly (but not) daily guarantees often benefi t more from agency- developed rosters and should avoid cafeteria rostering if possible. Tip

Chapter 6. Rostering 6-17 Level ͭ. Basic Rostering LEVEL 1 End of Basic Rostering. The Intermediate Section of Rostering continues on the next page. To jump to Rail Scheduling, go to page ͳ-ͭ.

Level ͮ. Intermediate RosteringChapter 6. Rostering 6-18 6.2 Intermediate Rostering The basic section of this chapter set the foundation for rostering using a relatively simple example. In this section, you will develop rosters for Route ͵ͳ based on the Run Guide from the intermediate section of Chapter ͱ: Runcutting (shown below). This Run Guide includes eight full-time runs, as shown below. Completed Run Cut - Blocks Rehooked Time Place Type Time Place Type 101 Str 1 5:31 5:46 Garage Pull 13:09 A Street 13:34 7:23 0:30 0:10 8:03 8:03 0:00 0:01:30 0:00:00 8:04:30 1.094 102 Str 1 12:44 13:09 A Street 19:53 Garage Pull 20:08 6:44 0:30 0:10 7:24 7:24 0:36 0:00:00 0:00:00 8:00:00 1.188 103 Str 2 5:35 5:50 Garage Pull 12:39 A Street 13:04 6:49 0:30 0:10 7:29 7:29 0:31 0:00:00 0:00:00 8:00:00 1.174 104 Str 2 12:14 12:39 A Street 19:49 Garage Pull 20:04 7:10 0:30 0:10 7:50 7:50 0:10 0:00:00 0:00:00 8:00:00 1.116 105 Str 3 5:46 6:01 Garage Pull 12:09 A Street 12:34 6:08 0:30 0:10 6:48 6:48 1:12 0:00:00 0:00:00 8:00:00 1.304 106 Str 3 11:44 12:09 A Street 19:19 Garage Pull 19:34 7:10 0:30 0:10 7:50 7:50 0:10 0:00:00 0:00:00 8:00:00 1.116 4 5:50 6:05 Garage Pull 10:19 Garage Pull 10:24 6 14:31 14:46 Garage Pull 18:19 Garage Pull 18:34 5 6:05 6:20 Garage Pull 9:53 Garage Pull 9:58 7 14:50 15:05 Garage Pull 18:38 Garage Pull 18:53 56:17 4:40 0:00 1:00 61:57 2:43 0:20 2:46 67:46 1.204 Run # Type Block # Report Time Start End Spread Penalty Paid Hours Pay/ Plat Sign Off Time Plat Hours Report Hours Paid Break Travel Hours Total Work Hours 8:37 12:44 0:00 Spread Make-up Overtime 0:18:30107 Spl 7:47 0:50 0:00 1:22:00 10:17:30 1.322 108 Spl 7:06 0:50 0:00 1.3247:56 12:48 0:04 0:00:00 1:24:00 9:24:00 Cafeteria-style rosters will be considered fi rst, since these are simpler for the scheduler. In a cafeteria system, the scheduler is only responsible for calculating the number of operators needed and the available days off by day of the week. A discussion of agency-developed ros- ters follows the cafeteria rostering. Cafeteria Rostering—Weekday-only Service The Run Guide for Route ͵ͳ, as shown above, has a total of eight daily runs. In this cafeteria rostering example, assume that the route operates only on weekdays and there are no Ͱ/ͭͬ rosters. The formulas for calculating days off and the number of operators required are shown below, yielding the following result: LEVEL 2

Chapter 6. Rostering 6-19 Level ͮ. Intermediate Rostering OPERATORS REQUIRED FOR 8-HOUR RUNS Weekly total Day # daily runs # days per week (daily runs * days) Weekdays (M-F) 8 5 40 Saturdays 0 1 0 Sundays 0 1 0 Weekly Total 40 Total Operators (weekly total runs / 5 days of work per operator) 8 DISTRIBUTION OF DAYS OFF total # Operators off each day available operators # daily runs (# operators - # daily runs) Weekdays (M-F) 8 8 0 Saturdays 8 0 8 Sundays 8 0 8 Weekly Total (weekdays * 5 + Saturday + Sunday) 16 Operators have no choice but to take Saturday and Sunday as their days off . The most straightforward operating roster that could result from cafeteria-style selection of work assign- ments would include the same run for each operator Monday through Friday. An example of this weekly roster is shown on the following page, under the assumption that the most senior operators would pick weekly assignments with overtime. Of course, the scheduler cannot pre- dict how the cafeteria roster will look.

Level Í®. Intermediate RosteringChapter 6. Rostering 6-20 CAFETERIA ROSTER - Route 97 Roster No. Sun Mon Tue Wed Thu Fri Sat Weekly Hours Platform Hours 101 Off 07 07 07 07 07 Off 10:17 10:17 10:17 10:17 10:17 51:27:30 38:55:00 102 Off 08 08 08 08 08 Off 9:24 9:24 9:24 9:24 9:24 47:00:00 35:30:00 103 Off 01 01 01 01 01 Off 8:04 8:04 8:04 8:04 8:04 40:22:30 36:55:00 104 Off 05 05 05 05 05 Off 8:00 8:00 8:00 8:00 8:00 40:00:00 30:40:00 105 Off 02 02 02 02 02 Off 8:00 8:00 8:00 8:00 8:00 40:00:00 33:40:00 106 Off 03 03 03 03 03 Off 8:00 8:00 8:00 8:00 8:00 40:00:00 34:05:00 107 Off 04 04 04 04 04 Off 8:00 8:00 8:00 8:00 8:00 40:00:00 35:50:00 108 Off 06 06 06 06 06 Off 8:00 8:00 8:00 8:00 8:00 40:00:00 35:50:00 338:50:00 281:25:00TOTAL WEEKLY HOURS

Chapter 6. Rostering 6-21 Level ͮ. Intermediate Rostering Cafeteria Rostering—Weekday and Weekend Service In this example, Route ͵ͳ operates seven days a week. Weekend service is the same on Sat- urday and Sunday in this example, and is about Ͳͬ% of weekday service. The fi rst step is to determine the available days off and the number of operators needed. The master run list is shown below. MASTER RUN LIST Weekday Runs Saturday Runs Sunday Runs Run # Pay hours Run # Pay hours Run # Pay hours 01 8:04 601 8:00 701 8:00 02 8:00 602 8:00 702 8:00 03 8:00 603 8:00 703 8:00 04 8:00 604 8:30 704 8:30 05 8:00 605 9:51 705 9:51 06 8:00 07 10:17 08 9:24 The total number of operators is determined using the formula for computing operators and days off , as shown below. The addition of weekend service on Route ͵ͳ results in a total of ͭͬ operators needed for this example, compared to eight for weekday-only service. When Satur- day and Sunday runs are part of the cafeteria rostering process, drivers also need to pick their days off . Since only fi ve operators are needed for Saturday and Sunday service, and ͭͬ opera- tors are needed in total, there will be fi ve operators “off ” on Saturdays and Sundays. Eight operators are required each weekday, so two of the ͭͬ operators can choose to be off on each weekday. The table on the following page summarizes the expected rostering outcomes. This goes back to the approach we have discussed in previous sections—that is, to know the outcome before you work through the details and develop a solution.

Level ͮ. Intermediate RosteringChapter 6. Rostering 6-22 OPERATORS REQUIRED FOR 8-HOUR RUNS WITH WEEKEND SERVICE Weekly total Day # daily runs # days per week (daily runs * days) Weekdays (M-F) 8 5 40 Saturdays 5 1 5 Sundays 5 1 5 Weekly Total 50 Total Operators (weekly total runs / 5 days of work per operator) 10 DISTRIBUTION OF DAYS OFF Total # Operators off each day available operators # daily runs (# operators - # daily runs) Weekdays (M-F) 10 8 2 Saturdays 10 5 5 Sundays 10 5 5 Weekly Total (weekdays * 5 + Saturday + Sunday) 20 What might the cafeteria roster look like for this example? Senior operators often prefer to maximize their pay hours and to have weekends free. The nature of the run (straight vs. split) and run begin and end times are also factors considered by operators. Assume that the fi rst two operators maximize their work hours and pick Saturday and Sunday as their days off , the next two are willing to work one weekend day, and the next two are will- ing to work both weekend days. The cafeteria roster after these six operators pick their work would look like the example on the following page.

Chapter 6. Rostering 6-23 Level ͮ. Intermediate Rostering CAFETERIA ROSTER - Route 97 (7 days) Roster No. Sun Mon Tue Wed Thu Fri Sat Platform Hours Work Hours Make-up Time Overtime Spread Time W 1001 Off 107 107 107 107 107 Off 10:17 10:17 10:17 10:17 10:17 38:55:00 43:05:00 0:00 3:05:00 13:40 1002 Off 108 108 108 108 108 Off 9:24 9:24 9:24 9:24 9:24 35:30:00 39:40:00 0:20 0:00:00 14:00 1003 Off Off 101 101 101 101 605 8:04 8:04 8:04 8:04 9:51 38:16:00 41:26:00 0:00 1:26:00 0:00 1004 705 105 105 105 105 Off Off 9:51 8:00 8:00 8:00 8:00 33:16:00 36:30:00 4:44 1:14:00 0:00 1005 704 101 Off Off 102 102 604 8:30 8:04 8:00 8:00 8:30 36:31:00 39:31:00 1:12 0:43:00 0:00 1006 Off 103 103 103 103 103 Off 8:00 8:00 8:00 8:00 8:00 34:05:00 37:25:00 2:35 0:00:00 0:00 Remaining days off available would be: DAYS OFF ALLOWED Updated Sunday Monday Tuesday Wednesday Thursday Friday Saturday 1 1 1 1 2 1 1 The completed cafeteria roster (shown below) has a weekly total of Ͱͮͯ:ͯͮ hours. This is ap- proximately ʹͰ hours more than the roster for weekday-only service, refl ecting the additional Ͱͮ hours of work on Saturday and Sunday. Most of this time (ͯ͵ͯ:ͮͰ work hours plus ͭͯ:ͬͰ make-up time) will be paid as straight time. Overtime accounts for Ͳ:ͮʹ hours and spread for ͮͳ:Ͱͬ. This assumes that make-up and overtime is calculated on a daily basis. With overtime and spread paid at time and a half and make-up time paid at the straight rate, the actual pay hours will be: Pay hours including make-up time, overtime, and spread penalty = ͯ͵ͯ:ͮͰ + ͭͯ:ͬͰ + (Ͳ:ͮʹxͬ.ͱ) + (ͮͳ:Ͱͬxͬ.ͱ) = Ͱͮͯ:ͯͮ

Level ͮ. Intermediate RosteringChapter 6. Rostering 6-24 CAFETERIA ROSTER - Route 97 (7 days) Roster No. Sun Mon Tue Wed Thu Fri Sat Platform Hours Work Hours Make-up Time Overtime Spread Time Weekly Pay Hours 1001 Off 107 107 107 107 107 Off 10:17 10:17 10:17 10:17 10:17 38:55:00 43:05:00 0:00 3:05:00 13:40 51:27:30 1002 Off 108 108 108 108 108 Off 9:24 9:24 9:24 9:24 9:24 35:30:00 39:40:00 0:20 0:00:00 14:00 47:00:00 1003 Off Off 101 101 101 101 605 8:04 8:04 8:04 8:04 9:51 38:16:00 41:26:00 0:00 1:26:00 0:00 42:09:00 1004 705 105 105 105 105 Off Off 9:51 8:00 8:00 8:00 8:00 33:16:00 36:30:00 4:44 1:14:00 0:00 41:51:00 1005 704 101 Off Off 102 102 604 8:30 8:04 8:00 8:00 8:30 36:31:00 39:31:00 1:12 0:43:00 0:00 41:04:30 1006 Off 103 103 103 103 103 Off 8:00 8:00 8:00 8:00 8:00 34:05:00 37:25:00 2:35 0:00:00 0:00 40:00:00 1007 Off 104 104 104 104 104 Off 8:00 8:00 8:00 8:00 8:00 35:50:00 39:10:00 0:50 0:00:00 0:00 40:00:00 1008 701 106 106 106 Off Off 601 8:00 8:00 8:00 8:00 8:00 36:30:00 39:30:00 0:30 0:00:00 0:00 40:00:00 1009 702 Off Off 102 106 106 602 8:00 8:00 8:00 8:00 8:00 36:04:00 39:04:00 0:56 0:00:00 0:00 40:00:00 1010 703 102 102 Off Off 105 603 8:00 8:00 8:00 8:00 8:00 34:36:00 38:03:00 1:57 0:00:00 0:00 40:00:00 359:33:00 393:24:00 13:04:00 6:28:00 27:40:00 423:32:00TOTAL It may be useful at this point to walk through an example of the calculations for a specifi c ros- ter number. In Roster ͭͬͬͭ, the daily platform hours for Run ͭͬͳ, taken from the Run Guide at the beginning of this section, are multiplied by ͱ to convert to weekly hours: ͳ:Ͱͳ x ͱ = ͯʹ:ͱͱ. Total hours are calculated by multiplying the daily total hours from the Run Guide for Run ͭͬͳ by ͱ: ʹ:ͯͳ x ͱ = Ͱͯ:ͬͱ. Make-up time is calculated in the same way, but Roster ͭͬͬͭ does not include any make-up time. Overtime was shown in the Run Guide as overtime pay hours (equal to one-half the overtime hours worked); the roster shows actual weekly overtime worked for clarity. Roster ͭͬͬͭ includ- ed ͯͳ minutes of daily overtime, which equals ͯ:ͬͱ weekly overtime. Spread time was also shown in the Run Guide as spread penalty (equal to one-half of actual spread hours); the roster shows actual weekly spread time for clarity. Run ͭͬͳ included ͮ:ͰͰ of daily spread time, which equals ͭͯ:Ͱͬ weekly spread time. Some operators calcu- late overtime on a weekly basis, while others calculate it on a daily basis, paying overtime for all hours over eight in a given day. To calculate the cost of your service, you will need to understand the policies that apply to your system. Calculating pay hours for a system that pays overtime daily requires making a calculation for each day of the week, and multiplying that day times the number of days that condition occurs in the year. Tip When calculating an- nual costs BE CARE- FUL—holidays that occur during the week often use weekend or special schedules. Do not assume that “Monday” service occurs on every Monday during a calendar year. Tip

Chapter 6. Rostering 6-25 Level ͮ. Intermediate Rostering The formula for weekly pay hours is: Work hours + Make-up time + (Overtime x ͬ.ͱ) + (Spread time x ͬ.ͱ) = Weekly pay hours Weekly pay hours For Roster ͭͬͬͭ = Ͱͯ:ͬͱ + ͬ + (ͯ:ͬͱ x ͬ.ͱ) + (ͭͯ:Ͱͬ x ͬ.ͱ) = ͱͭ:ͮͳ:ͯͬ The pay-to-platform ratio is one of the most useful means of evaluating rosters. In this exam- ple, the pay-to-platform ratio is: Pay to platform = Ͱͮͯ:ͯͮ / ͯͱ͵:ͯͯ = ͭ.ͭͳʹ Combining daily runs with signifi cant overtime with daily runs under eight hours can in- crease effi ciency when overtime and guarantees are calculated weekly. Tip

Level ͮ. Intermediate RosteringChapter 6. Rostering 6-26 Agency-developed Rostering—Weekday and Weekend Service With agency-developed rostering, the agency “pre-packages” the runs into weekly rosters. The operators then pick from the agency-developed list of weekly rosters. A number of factors infl uence how packages (Lines of Work) are developed for the weekly ros- ters. In cases where overtime and make-up time are calculated on a weekly and not daily basis, there is the opportunity to combine daily runs with greater than eight hours pay time with daily runs paying less than eight hours to reduce or eliminate weekly overtime and make-up time. In the basic rostering section, the agency-developed roster reduced the pay-to-platform ratio from ͭ.ͮͱʹ for the cafeteria roster to ͭ.ͮͮͱ. As shown earlier on the Master Run list, there are two long runs on weekdays and two on weekends. Several weekday runs (notably runs ͭͬͮ, ͭͬͯ, and ͭͬͱ) require at least ͯͬ minutes of make-up time daily. The objective of developing weekly rosters is to combine these runs into weekly work assignments that minimize overtime and make-up time. Ideally, work assign- ments will be as close to Ͱͬ hours per week as possible. A common approach to developing agency rosters is to adjust assignments with the most overtime or make-up time fi rst, and then address remaining assignments. Rosters ͭͬͬͭ and ͭͬͯ have the most weekly overtime, while Rosters ͭͬͬͰ, ͭͬͬͲ, ͭͬͭͬ, and ͭͬͬͱ have the most make-up time. In the example on the following page, switches in assignments are highlighted in bold. The fi rst switch is between Rosters ͭͬͬͭ and ͭͬͬͲ: the Wednesday through Friday runs have been exchanged to reduce overtime on Roster ͭͬͬͭ and make-up time on Roster ͭͬͬͲ. Next, the Saturday runs have been switched between Rosters ͭͬͬͯ and ͭͬͬʹ. The Tuesday runs on Rosters ͭͬͬͯ and ͭͬͬͰ have been exchanged. Finally, Run ͭͬͳ on Thursday, which had been switched from Roster ͭͬͬͭ to Roster ͭͬͬͲ, has been exchanged again with Run ͭͬͱ from Ros- ter ͭͬͬͰ. What is the end result of this agency-developed roster? See below—weekly overtime has been eliminated, and weekly make-up time has been reduced from ͭͯ:ͬͰ to ͳ:ͯͮ. Pay hours are now calculated as: Pay hours including make-up time, overtime, and spread penalty = ͯ͵ͮ:ͮʹ + ͳ:ͯͮ + (ͬ:ͬͬxͬ.ͱ) + (ͮͳ:Ͱͬxͬ.ͱ) = Ͱͭͯ:ͱͬ Pay to platform = Ͱͭͯ:ͱͬ / ͯͱ͵:ͯͯ = ͭ.ͭͱͭ

Chapter 6. Rostering 6-27 Level ͮ. Intermediate Rostering AGENCY-DEVELOPED ROSTER - Route 97 (7 days) Roster No. Sun Mon Tue Wed Thu Fri Sat Platform Hours Straight Time Make-up Time Overtime Spread Time Weekly Pay Hours 1001 Off 07 07 03 03 03 Off 10:17 10:17 8:00 8:00 8:00 36:01:00 39:41:00 0:19 0:00:00 5:28 42:44:00 1002 Off 08 08 08 08 08 Off 9:24 9:24 9:24 9:24 9:24 35:30:00 39:40:00 0:20 0:00:00 14:00 47:00:00 1003 Off Off 05 01 01 01 603 8:00 8:04 8:04 8:04 8:00 35:47:00 38:58:00 1:02 0:00:00 0:00 40:00:00 1004 705 05 01 05 07 Off Off 9:51 8:00 8:04 8:00 10:17 36:10:00 39:02:00 0:58 0:00:00 2:44 41:22:00 1005 704 01 Off Off 02 02 604 8:30 8:04 8:00 8:00 8:30 36:31:00 39:31:00 0:29 0:00:00 0:00 40:00:00 1006 Off 03 03 07 05 07 Off 8:00 8:00 10:17 8:00 10:17 35:20:00 39:01:00 0:59 0:00:00 5:28 42:44:00 1007 Off 04 04 04 04 04 Off 8:00 8:00 8:00 8:00 8:00 35:50:00 39:10:00 0:50 0:00:00 0:00 40:00:00 1008 701 06 06 06 Off Off 601 8:00 8:00 8:00 8:00 8:00 36:30:00 39:30:00 0:30 0:00:00 0:00 40:00:00 1009 702 Off Off 02 06 06 602 8:00 8:00 8:00 8:00 8:00 36:04:00 39:04:00 0:56 0:00:00 0:00 40:00:00 1010 703 02 02 Off Off 05 605 8:00 8:00 8:00 8:00 9:51 35:50:00 38:51:00 1:09 0:00:00 0:00 40:00:00 359:33:00 392:28:00 7:32:00 0:00:00 27:40:00 413:50:00TOTAL WEEKLY HOURS Weekly pay hours have been reduced from Ͱͮͯ:ͯͮ to Ͱͭͯ:ͱͬ. The agency-developed roster has reduced the pay-to-platform ratio from ͭ.ͭͳʹ to ͭ.ͭͱͭ. Typically, agency-developed rosters are more effi cient because of the ability to package long and short runs in a weekly roster. The reduction in pay-to-platform ratio was slight in this example only because the original roster was very effi cient. Another key benefi t is the ability to package a greater proportion of attractive Lines of Work. The outcome can be that the less senior employees are not just operating the undesirable runs or workweeks, and have an overall better working life. This has the potential to translate into lower staff turnover among the less senior employees, which is an issue for many transit sys- tems. One other consideration is that early and late runs should generally not be mixed in a given roster. The reasons for this are both fatigue-related (obvious if moving from a late fi nish to an early start), and also because, as a general rule, operators prefer low variance in the start and/or end times of their working days. The agency may have rules governing time off between

Level ͮ. Intermediate RosteringChapter 6. Rostering 6-28 shifts, but even without such rules it is not a good idea to have an operator pull in late in the evening and pull out early the following morning. A roster with early runs and a late run on the last work day would be acceptable under time-off rules, but operators generally do not fi nd this desirable. A comparison of the cafeteria and agency rostering results is shown below. COMPARISON CHART FOR CAFETERIA VS. AGENCY ROSTERING Cafeteria Agency Total hours for all weekly rosters 423:32:00 423:32:00 @ Straight time without make-up 393:24:00 392:28:00 @ Make-up hours 13:04:00 7:32:00 @ Overtime hours 6:28:00 0:00:00 @ Spread hours 27:40:00 27:40:00 Pay hours 423:32:00 413:50:00 Number of operators required 10 10 Total platform hours 359:33:00 359:33:00 Pay-to-platform ratio 1.178 1.151 Day Off Patterns Quality of working weeks is also an important consideration when developing rosters. This applies particularly to how day off patterns are developed and rest times are ensured (above minimum requirements). The scheduler should be able to predict the day off patterns before developing the rosters. For any fi xed set of input runs (i.e., the weekday, Saturday & Sunday runs to be rostered) there will be an ‘optimum’ set of day off patterns. Some of the consider- ations of day off patterns will include: Whether days off must be consecutive (particularly applies to four-day workweeks where two consecutive plus one other day may be acceptable) Which combinations of days are preferred. For example, is a Thursday/Friday combina- tion preferable to a Friday/Saturday combination (probably not)? Whether having an entire weekend off is an important consideration. If so more Thurs- day/Friday and Monday/Tuesday combinations will be preferred over Friday/Saturday and Sunday/Monday combinations, which will in turn allow more Saturday/Sunday weekend day off combinations to be created. The following graphics illustrate day off patterns for two options. Option ͭ requires consecutive days off . Option ͮ allows this to be traded off against additional full weekends off . The patterns that result are shown below: • • •

Chapter 6. Rostering 6-29 Level ͮ. Intermediate Rostering Combination Option 1 Option 2 Saturday/Sunday 4 5 Sunday/Monday 1 Monday/Tuesday 1 2 Tuesday/Wednesday 1 Wednesday/Thursday 1 1 Thursday/Friday 1 1 Friday/Saturday 1 Wednesday/Friday (Split Days) 1 As the table shows, revising day off combinations can achieve an additional entire weekend off (increases from four to fi ve) with the same set of input runs. However, this outcome requires a split days off combination, which may or may not be permissible under your labor rules. If split days off are not allowed, this would not be a valid option; however the example is provided to illustrate the concept of alternative approaches to days off patterns. Days off can be easily displayed as a basic roster, without the daily hours or numbers. This al- lows a more simple visualization of the roster pattern. In developing a roster it may be best to develop this blank pattern fi rst, and then simply “fi ll in the blanks” to complete the roster.

Level Í®. Intermediate RosteringChapter 6. Rostering 6-30 Roster Sun Mon Tue Wed Thur Fri Sat 1 Off Off 2 Off Off 3 Off Off 4 Off Off 5 Off Off 6 Off Off 7 Off Off 8 Off Off 9 Off Off 10 Off Off Days Off Pattern - Option 1 Roster Sun Mon Tue Wed Thur Fri Sat 1 Off Off 2 Off Off 3 Off Off 4 Off Off 5 Off Off 6 Off Off 7 Off Off 8 Off Off 9 Off Off 10 Off Off Days Off Pattern - Option 2

Chapter 6. Rostering 6-31 Level ͮ. Intermediate Rostering Agency-developed Rosters with Part-time Operators Part-time operators are allowed at many transit agencies, often with restrictions on the maxi- mum number of hours allowed per week or the number or percentage of part-timers. The examples in this section did not involve part-time operators because the runcut did not pro- duce any runs shorter than Ͳ:Ͱ͵. If your runcut results in runs of six hours or less, these can be combined into a weekly part-time roster (ͯͬ hours per week is a typical upper limit for part- time operators). Alternately, short and long runs might be mixed as in the preceding example to produce a more effi cient roster. The naming of which runs are part time and which are full time can be reconsidered as part of the rostering process, again depending upon what the labor agreement allows. This is dis- cussed in greater detail in the advanced section. Agency-developed Four-day and Five-day Rosters The development of four-day rosters is a viable option for many agencies. The ability to con- vert overtime to straight time is a primary advantage of four-day rosters (especially if overtime is paid on a daily basis), and is usually present if there are runs close to ͭͬ hours in length. The example in this section includes one weekday run longer than ͭͬ:ͬͬ, another weekday run ͵:ͮͰ in length, and a Saturday and Sunday run ͵:ͱͭ in length. Since each weekday run is operated fi ve days a week, these ͭͮ runs could have formed three four-day rosters. The remaining ͯʹ runs could then form seven fi ve-day rosters and one part-time roster consisting of three days’ work. However, the ability to mix and match runs in rosters would be limited if the longer runs are reserved for four-day rosters. A signifi cant portion of any cost savings from the use of four-day rosters may be off set by the added cost of hiring and training additional operators and paying additional fringe benefi ts. The iterative nature of runcutting and rostering may result in a diff erent runcut that would be conducive to four-day rosters without increasing the need for added operators. For many op- erators, a four-day roster is a welcome option because it provides an additional day off during the week. In short the number of operators required to operate four ͭͬ-hour runs over a workweek is usu- ally the same as required to operate fi ve eight-hour runs. While the ͭͬ-hour runs result in fewer runs at the runcut stage, the additional days off required for the four-day workweek result in the same number of total operators. In simple terms the ͮͱ% saving in runs (eight hours to ͭͬ hours) is off set by the equivalent additional days off required (ͮͱ% or one in four) by four-day workweeks.

Chapter 6. Rostering 6-32 LEVEL 2 End of Intermediate Rostering. The Advanced Section of Rostering continues on the next page. To jump to Rail Scheduling, go to page ͳ-ͭ. Level ͮ. Intermediate Rostering

Chapter 6. Rostering 6-33 Level ͯ. Advanced Rostering 6.3 Advanced Rostering The advanced section in Chapter ͱ: Runcutting did not develop any signifi cantly diff erent runcuts. Previous sections of this chapter have addressed the principal issues and approaches in rostering. Agencies that use cafeteria rostering have no control over the fi nal rosters, other than ensuring that they are in accord with all applicable rules and regulations. The operators put together their own rosters by selecting runs and days off from the master lists. Agencies that develop their own rosters may fi nd several opportunities to minimize costs: If make-up time and/or overtime are calculated on a weekly (Ͱͬ hours per week) rather than a daily (eight hours per day) basis, the scheduler can combine short and long runs in a given roster to avoid make-up time and overtime. As seen in the intermediate sec- tion, this strategy can reduce weekly pay hours and improve the pay-to-platform ratio. If make-up time and overtime are calculated on a daily basis, then mixing and matching will not achieve any savings. The tools available to a scheduler in these circumstances include whether and how much to utilize Ͱ/ͭͬ rosters and how to roster tripper runs. Ͱ/ͭͬ runs may generate savings by reducing the number of daily runs and the associ- ated travel, sign-in, and clear times. As noted in the advanced runcutting section, Ͱ/ͭͬ rosters will not reduce the number of operators needed, since the total number of op- erators required is a function of how much work is undertaken in a week, not in a given runcut. If there are many long runs around ͭͬ hours in length and if your agency allows four-day workweeks with ͭͬ-hour days, the scheduler should experiment with diff erent levels of Ͱ/ͭͬ rosters to assess the impacts. Some agencies even develop rosters with fi ve ͭͬ-hour daily runs as a method of reduc- ing the number of required operators. However, as expected, this rostering practice results in a high level of weekly overtime. See the discussion of overtime optimization in the advanced section of Chapter ͱ:Runcutting for a more in-depth discussion of po- tential benefi ts and pitfalls of high levels of overtime. Tripper runs may be packaged into part-time rosters if an agency uses part-time opera- tors. Alternately, agencies may choose not to assign these runs in a roster, but instead to leave them open to be operated by the extraboard. The runcut develops tripper runs, but these are not assigned an “operator type” until the rostering stage. The scheduler should be aware of the rostering and bidding impacts of the runs, and the rules that apply to part-time operators in selecting the runs that have been created. • • • • • LEVEL 3 tripper A short piece of work whose total time is less than that specifi ed as constituting a full-time run. A tripper is often a piece of work in the AM or PM peak period that cannot be combined with another piece of work to form a split run because of insuffi cient hours, excessive swing time, or excessive spread time. Trippers are often operated by extraboard or part-time operators. Tripper can also refer to a vehicle that pulls out, makes no more than one round-trip, and pulls in.

Level ͯ. Advanced RosteringChapter 6. Rostering 6-34 In some cases, the labor agreement may specify that a single roster can consist only of runs on the same route, or the same run type, or the same equipment type. Meeting these require- ments often results in separately identifi ed rosters for picking, e.g., regular and relief. Rules like these can also aff ect cafeteria rosters. For example, a modifi ed cafeteria roster may re- quire an operator to pick the same run on each weekday that it is available. The “quality” of agency-developed rosters is an important consideration. Rosters dictate the operator’s work life. When mixing and matching runs, care should be taken to combine AM runs only or PM runs only in a single roster. Operators also prefer an earlier end time on the day preceding a day off , although this cannot be managed for all rosters. Attention to opera- tor preferences when developing rosters can improve morale, often with no impact on costs or effi ciency. Advanced Topics in Rostering Revisiting Run Types Thus far we have kept run types as per the runcut—full time into a full-time roster, part time kept separate (or even packaged into part-time Lines of Work), and even kept ͱ/ʹ and Ͱ/ͭͬ ros- ters separate. However the rostering process has the capacity to revisit even basic assumptions about what a part-time run or full-time run is. Often a full-time or part-time run is determined not by daily hours but by weekly hours (this is not always the case, e.g., when the daily runs that can be assigned to an operator type are specifi ed). Therefore it is not the ʹ-hour day but the Ͱͬ-hour week that designates a full-time run. A part-time run, in some circumstances, may therefore be made up of three ͭͬ-hour runs, or four ͳ.ͱ-hour runs. These runs would normally be assumed as full time and at the runcut stage would have been created and designated as such. What would the benefi ts of such an approach be? One would be better and more reliable work hours/days for part-time employees. Consider a part-time employee that covered two week- end ͭͬ-hour runs and one additional weekday run. That would leave four days free to study, seek other employment, or for whatever lifestyle choice is preferred (often the reason for seek- ing a part-time job in the fi rst place). In this example one weekday run, one Saturday run, and one Sunday run have been covered by two employees, and with no overtime or guarantee costs (assuming weekly overtime and packaged Lines of Work). In addition, the potentially lower pay rates and benefi ts for part-time staff mean that in many cases maximum utilization of part-timers is effi cient—the issue often

Chapter 6. Rostering 6-35 Level ͯ. Advanced Rostering is that the types of work weeks off ered to part-time staff are not attractive, often being small peak pieces of work, or two small peak pieces of work separated by a long break. The aim of this section is not to attempt a prescriptive solution, as the “optimal” outcome will depend on a range of specifi c labor rules and conditions. Some of these will include: Daily and weekly hour limitations Pay rates and benefi ts for full-time and part-time staff Peak-to-base ratio, which may mandate a signifi cant number of shorter peak runs Employee preferences The amount of weekend service operated Preferences for Ͱ/ͭͬ or ͱ/ʹ work weeks of full-time staff Capacity to attract and retain part-time staff The intent here is to note the possibility to think beyond traditional scheduling approaches when rostering, and to reconsider how part-time and full-time staff are applied, and how run types can be redefi ned. Rotary Rostering The concept of a Rotating Roster was mentioned previously. In this type of roster, the opera- tors “rotate” or move from one Line of Work to the next. The concept is relatively simple in that an operator moves from one workweek to the next, throughout the sign-up period. This is depicted in the example below, based upon our Line ͵ͳ roster from the previous section. • • • • • • •

Level ͯ. Advanced RosteringChapter 6. Rostering 6-36 Operator Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9 Week 10 1 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 2 1002 1003 1004 1005 1006 1007 1008 1009 1010 1001 3 1003 1004 1005 1006 1007 1008 1009 1010 1001 1002 4 1004 1005 1006 1007 1008 1009 1010 1001 1002 1003 5 1005 1006 1007 1008 1009 1010 1001 1002 1003 1004 6 1006 1007 1008 1009 1010 1001 1002 1003 1004 1005 7 1007 1008 1009 1010 1001 1002 1003 1004 1005 1006 8 1008 1009 1010 1001 1002 1003 1004 1005 1006 1007 9 1009 1010 1001 1002 1003 1004 1005 1006 1007 1008 10 1010 1001 1002 1003 1004 1005 1006 1007 1008 1009 Week Roster Sun Mon Tue Wed Thu Fri Sat 704 01 02 02 604 8:30 8:04 8:00 8:00 8:30 03 03 07 05 07 8:00 8:00 10:17 8:00 10:17 04 04 04 04 04 8:00 8:00 8:00 8:00 8:00 701 06 06 06 601 8:00 8:00 8:00 8:00 8:00 702 02 06 06 602 8:00 8:00 8:00 8:00 8:00 703 02 02 05 605 8:00 8:00 8:00 8:00 9:51 07 07 03 03 03 10:17 10:17 8:00 8:00 8:00 08 08 08 08 08 9:24 9:24 9:24 9:24 9:24 05 01 01 01 603 8:00 8:04 8:04 8:04 8:00 705 05 01 05 07 9:51 8:00 8:04 8:00 10:17 Off Off Off Off Off ROTATION - OPERATOR 5 Off Off Off Off Off Off Off1009 1010 Off Off Off Off Off Off Off Off 5 6 1001 1002 1003 1004 1005 1006 1007 1008 7 8 9 10 1 2 3 4

Chapter 6. Rostering 6-37 Level ͯ. Advanced Rostering Some complexity can appear when moving from one workweek to the next. In the example above the operator moves from Roster Line ͭͬͬͱ to ͭͬͬͲ. However, that results in only one day off between Thursday in week one and Friday in week two—this may not meet labor agree- ment requirements. The consequence is that the roster needs to be built with regard to meeting requirements for the entire rotating period—so that changes as operators move through the diff erent packages from week to week fall within labor agreement requirements. This can signifi cantly complicate the rostering process. The outcome of this approach, used predominantly outside North America, is that the work- weeks are shared equally by the operators over the roster period. In order to mitigate signifi - cant lifestyle changes from week to week (e.g., moving from split runs one week to AM runs the next week to PM runs the following week) it may be that several distinct rosters are created (e.g., a “Split runs” roster, an “AM Runs” roster, a “PM Runs” roster, etc.), and operators rotate within those distinct rosters. Rostering for Longer Periods While the base rostering concept typically covers a repeating workweek there are situations where rosters are built for longer periods. Typically this will be either four or ͭͮ weeks, but may vary. For longer periods the process and issue are very much the same as for the weekly roster. However, the Rotating Roster concept above noted the issue of meeting labor agreement re- quirements over more than one week. This can apply to a range of labor agreement rules. The kinds of considerations that need to be monitored include: Days off requirements over ͭ week and multiple weeks Rest times, particularly when moving from one week to the next Total work hours over one week and the entire period (which may also be limited by driving time regulations) Rostering for longer periods adds additional levels of complexity to the rostering process. Below is an example of what part of the roster may look like in our Line ͵ͳ example. Note the similar workweeks to the weekly roster, but days off patterns are adjusted to ensure legality. • • •

Level ͯ. Advanced RosteringChapter 6. Rostering 6-38 Roster Week Sun Mon Tue Wed Thu Fri Sat 704 101 102 102 604 8:30 8:04 8:00 8:00 8:30 105 101 101 101 603 8:00 8:04 8:04 8:04 8:00 103 103 107 105 107 8:00 8:00 10:17 8:00 10:17 104 104 104 104 104 8:00 8:00 8:00 8:00 8:00 701 106 106 106 601 8:00 8:00 8:00 8:00 8:00 702 102 106 106 602 8:00 8:00 8:00 8:00 8:00 703 102 102 105 605 8:00 8:00 8:00 8:00 9:51 107 107 103 103 103 10:17 10:17 8:00 8:00 8:00 Off Off Off Off Off 4-Week Roster Off Off Off Off Off Off 1001 Off Off 4 Off Off 1002 1 2 3 4 1 2 3 Off A More Automated Spreadsheet We have discussed in previous chapters the need to utilize automated spreadsheet functional- ity to enhance spreadsheet-based scheduling. This applies strongly to rostering under either the cafeteria- or scheduler-developed approaches. We will reiterate a few key points: Limit the amount of typed data to those things that must be typed. Anything that can be a formula, should be a formula—without exception . This will reduce errors, minimize data input time, and signifi cantly enhance the process. Typed values in a spreadsheet are one of the greatest sources of errors in the spreadsheet-based scheduling process and are often unnecessary, i.e., they could have been calculated values. Use error trapping techniques. Simple checks can be developed to automatically iden- • •

Chapter 6. Rostering 6-39 Level ͯ. Advanced Rostering tify errors or alert the scheduler to issues. Use formatting to enhance the output and “viewability.” Spreadsheets can be format- ted to produce elegant reports for distribution, or to assist the scheduler in viewing solutions. Use conditional formatting to further enhance the presentation, for error identifi cation, or to highlight certain outcomes. Let’s look again at the basic ͭͬ-operator roster developed in the previous section. But this time we will only allow typed input into two areas—the daily runs and the basic roster information (which runs form a roster). These areas are shaded in dark gray below. ROSTER INPUT TABLE Sun Mon Tue Wed Thu Fri Sat 704 101 102 102 604 103 103 107 105 107 104 104 104 104 104 701 106 106 106 601 702 102 106 106 602 703 102 102 105 605 107 107 103 103 103 108 108 108 108 108 105 101 101 101 603 705 105 101 105 107 Sum 3,515 836 836 836 836 836 3,015 Check x x x x x x x ROSTER INPUT TABLE RUNS TO BE ROSTERED 101 8:04 601 8:00 701 8:00 102 8:00 602 8:00 702 8:00 103 8:00 603 8:00 703 8:00 104 8:00 604 8:30 704 8:30 105 8:00 605 9:51 705 9:51 106 8:00 3,015 42:21 3,515 42:21 107 10:17 108 9:24 836 67:45 Weekday Runs Saturday Runs Sunday Runs • •

Level ͯ. Advanced RosteringChapter 6. Rostering 6-40 These basic inputs are then used to generate the roster output, which looks very much like that presented in the previous section but without the totals (for simplicity sake, since the totals are easily calculated by basic formulas anyway). ROSTER OUTPUT Sun Mon Tue Wed Thu Fri Sat Total Hours 704 101 Off Off 102 102 604 8:30 8:04 8:00 8:00 8:30 41:04 Off 103 103 107 105 107 Off 8:00 8:00 10:17 8:00 10:17 44:34 Off 104 104 104 104 104 Off 8:00 8:00 8:00 8:00 8:00 40:00 701 106 106 106 Off Off 601 8:00 8:00 8:00 8:00 8:00 40:00 702 Off Off 102 106 106 602 8:00 8:00 8:00 8:00 8:00 40:00 703 102 102 Off Off 105 605 8:00 8:00 8:00 8:00 9:51 41:51 Off 107 107 103 103 103 Off 10:17 10:17 8:00 8:00 8:00 44:34 Off 108 108 108 108 108 108 9:24 9:24 9:24 9:24 9:24 47:00 Off Off 105 101 101 101 603 8:00 8:04 8:04 8:04 8:00 40:12 705 105 101 105 107 Off Off 9:51 8:00 8:04 8:00 10:17 44:12 Sum 42:21 67:45 67:45 67:45 67:45 67:45 42:21 423:27 Check x x x x x x x ROSTER OUTPUT Every cell in the above table is calculated through the use of lookup functions. The only typed value then becomes the daily inputs above, which can be manipulated and adjusted, and the results are seen immediately.

Chapter 6. Rostering 6-41 Level ͯ. Advanced Rostering Formulas are shown below. Only a subset of the output table is shown for readability purposes. Sun Mon Tue Wed Thu =IF(B7=0,"Off",B7) =IF(C7=0,"Off",C7) =IF(D7=0,"Off",D7) =IF(E7=0,"Off",E7) =IF(F7=0,"Off",F7) =VLOOKUP(B7,$S$7:$T$15,2) =VLOOKUP(C7,$M$7:$N$15,2) =VLOOKUP(D7,$M$7:$N$15,2) =VLOOKUP(E7,$M$7:$N$15,2) =VLOOKUP(F7,$M$7:$N$15,2) =IF(B8=0,"Off",B8) =IF(C8=0,"Off",C8) =IF(D8=0,"Off",D8) =IF(E8=0,"Off",E8) =IF(F8=0,"Off",F8) =VLOOKUP(B8,$S$7:$T$15,2) =VLOOKUP(C8,$M$7:$N$15,2) =VLOOKUP(D8,$M$7:$N$15,2) =VLOOKUP(E8,$M$7:$N$15,2) =VLOOKUP(F8,$M$7:$N$15,2) =IF(B9=0,"Off",B9) =IF(C9=0,"Off",C9) =IF(D9=0,"Off",D9) =IF(E9=0,"Off",E9) =IF(F9=0,"Off",F9) =VLOOKUP(B9,$S$7:$T$15,2) =VLOOKUP(C9,$M$7:$N$15,2) =VLOOKUP(D9,$M$7:$N$15,2) =VLOOKUP(E9,$M$7:$N$15,2) =VLOOKUP(F9,$M$7:$N$15,2) =IF(B10=0,"Off",B10) =IF(C10=0,"Off",C10) =IF(D10=0,"Off",D10) =IF(E10=0,"Off",E10) =IF(F10=0,"Off",F10) =VLOOKUP(B10,$S$7:$T$15,2) =VLOOKUP(C10,$M$7:$N$15,2) =VLOOKUP(D10,$M$7:$N$15,2) =VLOOKUP(E10,$M$7:$N$15,2) =VLOOKUP(F10,$M$7:$N$15,2) =IF(B11=0,"Off",B11) =IF(C11=0,"Off",C11) =IF(D11=0,"Off",D11) =IF(E11=0,"Off",E11) =IF(F11=0,"Off",F11) =VLOOKUP(B11,$S$7:$T$15,2) =VLOOKUP(C11,$M$7:$N$15,2) =VLOOKUP(D11,$M$7:$N$15,2) =VLOOKUP(E11,$M$7:$N$15,2) =VLOOKUP(F11,$M$7:$N$15,2) =IF(B12=0,"Off",B12) =IF(C12=0,"Off",C12) =IF(D12=0,"Off",D12) =IF(E12=0,"Off",E12) =IF(F12=0,"Off",F12) =VLOOKUP(B12,$S$7:$T$15,2) =VLOOKUP(C12,$M$7:$N$15,2) =VLOOKUP(D12,$M$7:$N$15,2) =VLOOKUP(E12,$M$7:$N$15,2) =VLOOKUP(F12,$M$7:$N$15,2) =IF(B13=0,"Off",B13) =IF(C13=0,"Off",C13) =IF(D13=0,"Off",D13) =IF(E13=0,"Off",E13) =IF(F13=0,"Off",F13) =VLOOKUP(B13,$S$7:$T$15,2) =VLOOKUP(C13,$M$7:$N$15,2) =VLOOKUP(D13,$M$7:$N$15,2) =VLOOKUP(E13,$M$7:$N$15,2) =VLOOKUP(F13,$M$7:$N$15,2) =IF(B14=0,"Off",B14) =IF(C14=0,"Off",C14) =IF(D14=0,"Off",D14) =IF(E14=0,"Off",E14) =IF(F14=0,"Off",F14) =VLOOKUP(B14,$S$7:$T$15,2) =VLOOKUP(C14,$M$7:$N$15,2) =VLOOKUP(D14,$M$7:$N$15,2) =VLOOKUP(E14,$M$7:$N$15,2) =VLOOKUP(F14,$M$7:$N$15,2) =IF(B15=0,"Off",B15) =IF(C15=0,"Off",C15) =IF(D15=0,"Off",D15) =IF(E15=0,"Off",E15) =IF(F15=0,"Off",F15) =VLOOKUP(B15,$S$7:$T$15,2) =VLOOKUP(C15,$M$7:$N$15,2) =VLOOKUP(D15,$M$7:$N$15,2) =VLOOKUP(E15,$M$7:$N$15,2) =VLOOKUP(F15,$M$7:$N$15,2) =IF(B16=0,"Off",B16) =IF(C16=0,"Off",C16) =IF(D16=0,"Off",D16) =IF(E16=0,"Off",E16) =IF(F16=0,"Off",F16) =VLOOKUP(B16,$S$7:$T$15,2) =VLOOKUP(C16,$M$7:$N$15,2) =VLOOKUP(D16,$M$7:$N$15,2) =VLOOKUP(E16,$M$7:$N$15,2) =VLOOKUP(F16,$M$7:$N$15,2) =B27+B29+B31+B33+B35+B37+B3=C27+C29+C31+C33+C35+C37+C3=D27+D29+D31+D33+D35+D37+D=E27+E29+E31+E33+E35+E37+E39=F27+F29+F31+F33+F35+F37+F39 =IF(B47=$T$13,"x","ERR") =IF(C47=$N$16,"x","ERR") =IF(D47=$N$16,"x","ERR") =IF(E47=$N$16,"x","ERR") =IF(F47=$N$16,"x","ERR") A further enhancement to the sheet is the checking and visualization of data through the use of conditional formatting. In the example below the weekly totals are shown according to a preferred range—less than the preferred range in italic, more than the preferred range in bold. Again this is automated, where the cell formatting references the cells with the preferred range.

Level ͯ. Advanced RosteringChapter 6. Rostering 6-42 Sun Mon Tue Wed Thu Fri Sat Total Hours 704 101 Off Off 102 102 604 8:30 7:04 8:00 8:00 8:30 40:04 Off 103 103 107 105 107 Off 8:00 8:00 10:17 8:00 10:17 44:34 Off 104 104 104 104 104 Off 8:00 8:00 8:00 8:00 8:00 40:00 701 106 106 106 Off Off 601 8:00 8:00 8:00 8:00 8:00 40:00 702 Off Off 102 106 106 602 8:00 8:00 8:00 8:00 8:00 40:00 703 102 102 Off Off 105 605 8:00 8:00 8:00 8:00 9:51 41:51 Off 107 107 103 103 103 Off 10:17 10:17 8:00 8:00 8:00 44:34 Off 108 108 108 108 108 108 9:24 9:24 9:24 9:24 9:24 47:00 Off Off 105 101 101 101 603 8:00 8:04 8:04 8:04 8:00 40:12 705 105 101 105 107 Off Off 9:51 8:00 8:04 8:00 10:17 44:12 Sum 42:21 66:45 67:45 67:45 67:45 67:45 42:21 422:27 Check x ERR x x x x x ROSTER OUTPUT Preferred Range Max ϦϤ:ϢϢ Min ϦϢ:ϥϢ Finally the above table shows an error check, based on a simple formula (that checks to be sure the total hours assigned for a day match the total daily run hours). The formula result is high- lighted with a conditional format. In the example above, an incorrect time was entered manu- ally for Run ͭͬͭ on Monday, leading to an error message in the row marked “Check.” This example provides an indication of some basic spreadsheet functionality that can be ap- plied to automate manual processes and reduce the risk of error. Current spreadsheets have

Chapter 6. Rostering 6-43 Level ͯ. Advanced Rostering signifi cant features, including the ability to write specifi c code, that enable a range of eff ective approaches to assist the scheduling process. Computerized/Automated Rostering Rostering is another part of the scheduling process to benefi t from computerized programs, in many of the typical ways. Rostering constantly involves making trade-off s between effi ciency, quality, and operator preferences. Often these preferences counteract each other and so a bal- ance must be found. In particular, the ability to run numerous iterations with almost instant results (due to the rela- tively simple math involved) allows the scheduler good control over solutions. The basic scheduling tenets discussed previously still hold true—you should know what the answer will be before the computer provides it, you must question the outputs, and you should be prepared to run numerous iterations until a fi nal solution is achieved. Day off patterns may or may not be developed automatically by the computer program. If so, great—but check and test it. If not, work through it yourself. The computer can then “fi ll in the empty boxes” of the blank pattern, subject to the hard and soft rules defi ned. Then the solution can be reviewed. And so it goes—refi ne the rules, run again, review again. ͭ. Hard rules. The types of hard rules will include factors such as: Minimum/maximum work hours in a period Rest break requirements Total days worked in a week or period Day off requirements (must they be consecutive, even for four-day weeks?) Other restrictions around run types, similar start/fi nish times, or routes operated ͮ. Constraints. As with runcutting there will be a series of constraints, or “soft” rules. These will tend to address the kinds of tradeoff s discussed throughout this section— weekends off versus split days off , start/fi nish time harmonization versus preferred work times. ͯ. Cost Elements. And of course there will be required costing inputs, including: Labor pay rates Overtime and guarantee requirements Benefi ts • • • • • • • •

Level ͯ. Advanced RosteringChapter 6. Rostering 6-44 The rostering function will tend to look very much like a spreadsheet—a pattern with cells for each work day, and each row representing a workweek. Typically the system will allow the scheduler to manually input or swap runs and immediately see the costing (or other impacts). As with all other aspects of scheduling the computerized software is only a tool at the disposal of the scheduler. In most cases a purely automated solution is unlikely to be the “best” solu- tion, as it is diffi cult to successfully capture all preferences/approaches in a form that a system can compute at a mathematical level. This is why running a number of iterations and working interactively with the system is necessary. Rostering Outputs The examples here have provided some basic outputs. In reality rosters tend to be larger and require additional information to be provided to downstream users. The kinds of reports to be utilized during the rostering should include: Statistics, inclusive of costs, manpower requirements, preferably broken down and ag- gregated as required. Outputs that display the roster in a simple format that provides the required informa- tion for various users (the scheduler, operations staff , operators, etc.). This could in- clude the run number, start/fi nish times and worked hours, across the week, for each roster assignment. Error checking reports to ensure all runs are assigned, all are legal, etc. The rostering reports will typically be customized to suit the requirements of a particular agency. One additional sample report is provided on the opposite page. We include it since it provides an indication of the volume of information that can be generated without great dif- fi culty and the amount of time that can be shown. In this case it provides the master roster for three operators for an entire year. Extraboard Considerations The topic of the extraboard has been raised but never fully discussed at various points through- out this manual. Extraboard considerations are presented here. Why mention extraboard in a scheduling manual? Well, because at many transit systems, schedule personnel are responsible for helping to “size” the extraboard and generally to sign up operators to staff it (or to “hold it down”). For those somewhat unfamiliar with the extraboard (generally called spareboard in Canada), it is a group of jobs that is fi lled at the time of the sign-up with operators who choose to be • • •

Chapter 6. Rostering 6-45 Level ͯ. Advanced Rostering JAN FEB MAR APR MAY JUN JUL AUG SEP 0CT N0V DEC DAY Roster 1 28 21 14 6 29 MON 03 6:15 15:15 9:00 06 5:05 15:05 10:00 04 14:50 22:50 8:00 29 22 15 7 30 TUE 04 14:50 22:50 8:00 03 6:15 15:15 9:00 30 23 16 8 31 WED 02 14:51 22:37 7:46 06 5:05 15:05 10:00 01 5:50 15:21 9:31 31 24 17 9 THU 02 14:51 22:37 7:46 01 5:50 15:21 9:31 1 25 18 10 FRI 04 14:50 22:50 8:00 03 6:15 15:15 9:00 2 26 19 11 SAT 01 5:30 15:00 9:30 3 27 20 12 SUN 02 8:418:40 17:21 WEEKLY HOURS 42:57 44:46 36:02 4 28 21 13 MON 01 5:50 15:21 9:31 02 14:51 22:37 7:46 5 29 22 14 TUE 01 5 50 15 21 9 31 02 14 51 22 37 7 46OFF OFF OFF OFF OFF OFF OFF OFF OFF H0URSH0URS Roster 3H0URS Roster 2 : : : : : : 6 30 23 15 WED 02 14:51 22:37 7:46 01 5:50 15:21 9:31 7 1 24 16 THU 02 14:51 22:37 7:46 01 5:50 15:21 9:31 8 2 25 17 FRI 02 14:51 22:37 7:46 01 5:50 15:21 9:31 06 5:05 17:05 12:00 9 3 26 18 SAT 06 11:00 21:00 10:00 01 5:30 15:00 9:30 10 4 27 19 SUN 02 8:418:40 17:21 WEEKLY HOURS 42:49 38:04 45:43 FORTNIGHTLY HOURS 85:46 82:50 81:45 11 5 28 20 12 6 29 21 13 7 30 22 14 8 31 23 15 9 1 24 16 10 2 25 17 11 3 26 WEEKLY HOURS 18 12 4 27 OFF OFF OFF OFFOFF 19 13 5 28 20 14 6 29 21 15 7 30 22 16 8 31 23 17 9 1 24 18 10 2 WEEKLY HOURS FORTNIGHTLY HOURS 25 19 11 3 26 20 12 4 27 21 13 5 28 22 14 6 29 23 15 7 1 24 16 8 2 25 17 9 WEEKLY HOURS 4-WEEK HOURS 3 26 18 10 4 27 19 11 5 28 20 12 6 29 21 13 7 30 22 14 8 31 23 15 9 1 24 16 WEEKLY HOURS FORTNIGHTLY HOURS 10 2 25 17 11 3 26 18 12 4 27 19 13 5 28 20 14 6 29 21 15 7 30 22 16 8 31 23 WEEKLY HOURS 17 9 1 24 18 10 2 25 19 11 3 26 20 12 4 27 21 13 5 28 22 14 6 29 23 15 7 30 WEEKLY HOURS FORTNIGHTLY HOURS 24 16 8 1 25 17 9 2 TUE 04 14:50 22:50 8:00 03 6:15 15:15 9:00 2 26 18 10 3 WED 02 14:51 22:37 7:46 06 5:05 15:05 10:00 01 5:50 15:21 9:31 3 27 19 11 4 THU 02 14:51 22:37 7:46 01 5:50 15:21 9:31 4 28 20 12 5 FRI 04 14:50 22:50 8:00 03 6:15 15:15 9:00 5 29 21 13 6 SAT 01 5:30 15:00 9:30 OFF OFF OFF OFF OFF 4-WEEK HOURS 6 30 22 14 7 SUN 02 8:418:40 17:21 WEEKLY HOURS 42:57 44:46 36:02 7 31 23 15 8 MON 01 5:50 15:21 9:31 OFF 02 14:51 22:37 7:46 8 1 24 16 9 TUE 01 5:50 15:21 9:31 02 14:51 22:37 7:46 9 2 25 17 10 WED 02 14:51 22:37 7:46 01 5:50 15:21 9:31 10 3 26 18 11 THU 02 14:51 22:37 7:46 01 5:50 15:21 9:31 11 4 27 19 12 FRI 02 14:51 22:37 7:46 01 5:50 15:21 9:31 06 5:05 17:05 12:00 12 5 28 20 13 SAT 06 11:00 21:00 10:00 01 5:30 15:00 9:30 13 6 29 21 14 SUN 02 8:418:40 17:21 WEEKLY HOURS 42:49 38:04 45:43 FORTNIGHTLY HOURS 85:46 82:50 81:45 14 7 30 22 15 MON 01 5:50 15:21 9:31 02 14:51 22:37 7:46 OFF 15 8 1 23 16 TUE 01 5:50 15:21 9:31 02 14:51 22:37 7:46 06 5:05 17:05 12:00 16 9 2 24 17 WED 06 12:00 23:00 11:00 02 14:51 22:37 7:46 01 5:50 15:21 9:31 17 10 3 25 18 THU 02 14:51 22:37 7:46 01 5:50 15:21 9:31 18 11 4 26 19 FRI 02 14:51 22:37 7:46 01 5:50 15:21 9:31 19 12 5 27 20 SAT 01 5:30 15:00 9:30 OFF OFF OFF OFF OFF OFF OFF OFF OFF OFF OFF OFF 20 13 6 28 21 SUN 02 8:418:40 17:21 WEEKLY HOURS 45:34 41:29 40:33 21 14 7 29 22 MON 01 5:50 15:21 9:31 02 14:51 22:37 7:46 22 15 8 30 23 TUE 01 5:50 15:21 9:31 02 14:51 22:37 7:46 23 16 9 1 24 WED 01 5:50 15:21 9:31 02 14:51 22:37 7:46 06 11:00 23:00 12:00 24 17 2 25 THU 01 5:50 15:21 9:31 02 14:51 22:37 7:46 25 3 26 FRI 01 5:50 15:21 9:31 02 14:51 22:37 7:46 26 4 27 SAT 06 10:00 21:00 9:30 2 28 SUN OFF OFF OFF OFFOFF OFF JAN FEB MAR APR MAY JUN JUL AUG SEP 0CT N0V DEC DAY Roster 1 R 28 21 14 6 29 MON 03 6:15 15:15 9:00 29 22 15 7 30 TUE 04 14:50 22:50 8:00 30 23 16 8 31 WED 02 14:51 22:37 7:46 31 24 17 9 THU 1 25 18 10 FRI 2 26 19 11 SAT 01 5:30 15:00 9:30 3 27 20 12 SUN 02 8:40 17:21 8:41 WEEKLY HOURS 42:57 4 28 21 13 MON 01 5:50 15:21 9:31 5 29 22 14 TUE OFF OFF OFF H0URS 6 30 23 15 WED 02 14:51 22:37 7:46 7 1 24 16 THU 02 14:51 22:37 7:46 8 2 25 17 FRI 02 14:51 22:37 7:46 9 3 26 18 SAT 06 11:00 21:00 10:00 10 4 27 19 SUN WEEKLY HOURS 42:49 FORTNIGHTLY HOURS 85:46 11 5 28 20 MON 01 5:50 15:21 9:31 12 6 29 21 TUE 01 5:50 15:21 9:31 13 7 30 22 WED 06 12:00 23:00 11:00 14 8 31 23 THU 02 14:51 22:37 7:46 15 9 1 24 FRI 02 14:51 22:37 7:46 16 10 2 25 SAT 17 11 3 26 SUN WEEKLY HOURS 45:34 18 12 4 27 MON OFF OFF OFF OFF 19 13 5 28 TUE 01 5:50 15:21 9:31 20 14 6 29 WED 01 5:50 15:21 9:31 21 15 7 30 THU 01 5:50 15:21 9:31 22 16 8 31 FRI 01 5:50 15:21 9:31 23 17 9 1 SAT 24 18 10 2 SUN WEEKLY HOURS 38:04 OFF OFF

Level ͯ. Advanced RosteringChapter 6. Rostering 6-46 “extra” instead of on a run, or are detailed there when there are no more runs to pick. Typi- cally new operators just out of training are placed on the extraboard. The board itself is used to cover work that would otherwise be open due to absence of the regularly assigned operator or because the work was not picked. Covering for absences is the primary job of the extraboard. Often extraboard operators are the newest operators with the least seniority. Operators who begin work at the property between sign-ups usually start on the extraboard because they do not yet have a permanent assignment. Many properties allow operators to select the extra- board with a fi xed report time and days off along with other runs during sign-ups. This tends to attract more senior operators who might prefer better days off or a fi nd a more desirable start time on the board. Many agencies rotate their extraboards. To understand this procedure, think of the extra- board as a collection of slots, each representing an operator. The slots move up the board to an earlier start time every day. The top slot “falls off ” and lands at the bottom. In this way operators working the board have the opportunity to work assignments at all times of the day, starting with the latest at the bottom and working up to the earliest at the top. These slots are matched up with open work which is sorted by Transportation dispatchers at each garage from the earliest to latest in starting time order. An agency may rotate the extraboard by one or more slots each day. The process of rotating the board helps “spread the wealth around” in giving each operator the opportunity to work the better (and often higher-paying) runs along with the less-desirable work. An added benefi t is that each operator falling to the bottom gets essentially most of a day off . Consider the op- erator at the top of the board today who draws an open run starting at Ͱ:ͬͬ AM and fi nishing around ͭ:ͬͬ PM. Tomorrow, he/she will be in the bottom slot and may have a Ͱ:ͬͬ PM report time. The work assigned to the board is varied. There are open runs, where the assigned operator might be on the sick list or may have retired. There are show ups, which are assignments to report at a certain time and stand by in case someone “misses out” or phones in sick at the last minute. There are vacation hold-downs, which cover for operators who have picked the week to use one or more of their vacation weeks. There are trippers—and this is where the schedul- ers come into the picture in a big way. Schedulers have control over the number and types of trippers that will be operated by the extraboard. The number of trippers for the AM and PM peaks must be either balanced (piece balancing) or arranged so they complement the other work that the extraboard has to cover. It is always wise for the schedulers to be in touch with each operating garage and understand their staffi ng situations and what is feasible in the way open run A run which was not bid at the pre- vious sign-up or has been vacated during a sign-up due to retirements, terminations, long term illnesses and such. piece balancing The process in which the scheduler strives to balance the number of AM and PM pieces to increase the possibility of creating split runs that are in accord with formal and infor- mal rules and to leave a balanced number of AM and PM trippers for the extraboard to cover Finding ways to encour- age more senior opera- tors to sign up for the extraboard is often in the best interest of the property, since these drivers may have a different assignment every day and work best when they have a working knowledge of the system. Tip

Chapter 6. Rostering 6-47 Level ͯ. Advanced Rostering of trippers that get left for them to fi ll. While not a common practice, at some agencies the labor agreement or garage staff give schedulers a target for the percentage of work left open. With regard to fi lling trippers, some agencies allow operators on regular runs to “bid” trip- pers during the sign-up. An operator having an AM run fi nishing by, say, ͭ:ͯͬ PM may choose to regularly work a PM tripper with the hours of ͯ:ͱͬ PM to Ͳ:ͭͬ PM. The tripper is worked at overtime, but the advantage to the agency of a regular operator steadily holding the work down is signifi cant. Not all trippers are biddable. In this case the tripper was small enough in hours so it does not put the operator at risk of (a) driving for more hours than a state or federal regulation would allow or (b) having too few hours off between assignments. Other work includes those show-ups, both AM and PM (although enough extraboard operators may be available who can cover both), special service (baseball extras and the like), charters (usually limited these days in the public transit sector so as not to compete with private trans- portation companies), doctor appointments, union business, and retraining. The list of reasons for absence is longer than this, to be sure, but it falls into two categories: known absences and unexpected absences. To eliminate as many surprises as possible, agencies require an operator who is sick to call in as soon as possible and get on the sick list. If it is before the time the next day’s extraboard is posted (usually around noon), his or her run is placed on the list of open runs to be fi lled. If the call comes after the next day’s extraboard is posted, the run is given to one of the show-up people. Vacation weeks off must be picked, usually at the sign-up closest to the fi rst of the year. Some systems set out special slots to be bid for vacation hold-downs and keep these separate from the extraboard, while others include vacations with other absences on the extraboard. At a number of systems, one week of vacation can be taken in individual days. These usually require at least a day of prior notifi cation to the dispatcher and may only be granted up to a certain number of operators on a given day. With all of these eff orts to “manage” absences, there are still unexpected no-shows, and for these the show-ups are slotted at times where they have been found through experience to be most helpful. If a show-up does not catch an assignment to fi ll in on a missing operator’s run, then he/she is usually kept on stand-by for a predetermined amount of time. If still not as- signed, the show-up is given a later report time or a piece of a run during the latter part of the work day. Dispatchers often get very creative in the use of their show-up operators. It is not unusual for an extraboard operator’s day to begin with a Ͱ:Ͱͱ AM show-up, then an assignment to work show-up An assignment for an extraboard operator to be at a specifi c location to fi ll in for a miss out or to do other work. A certain amount of stand-by time is paid in the event that the operator does not receive a run. stand-by time The time that an operator spends at the garage at the agency’s direction awaiting assignment of a run or a piece of work. Usually associated with a show-up by an extraboard operator to provide a pool of opera- tors that will be available to fi ll runs vacated by unscheduled absences.

Level ͯ. Advanced RosteringChapter 6. Rostering 6-48 an AM tripper from Ͳ:ͭͬ to ʹ:Ͱͱ AM, followed by a return to the drivers room on stand-by until released at ͭͭ:ͬͬ AM. That operator will likely be told to return on show-up at ͮ:ͬͬ PM and may catch a second half of a run where its operator has asked for early time off for personal busi- ness, or may work a PM tripper, fi nishing the day around Ͳ:ͬͬ PM. On some systems, there is a contractual maximum spread for the extraboard, just as there often is for regular runs. If our example property has a ͭͯ-hour spread maximum, then our operator would have to be given an assignment that would fi nish by ͱ:Ͱͱ PM. Extraboard Sizing The two main issues with managing an extraboard are (ͭ) proper sizing and (ͮ) assigning the board each day correctly, effi ciently, and without any appearance of favoritism. On systems with several garages, there is the added concern for consistency in how assigning is done across the system. With regard to the fi rst concern, how do we know just what the proper sizing is? Many systems simply go by a percentage that has worked in the past. In the scheduling survey, some agencies reported that their ideal extraboard size is ͭʹ%. That is, the number of operators on the extra- board is ͭʹ% of the number of operators needed to cover all scheduled runs and relief runs at that particular garage. For example, if the total number of runs to be picked at that garage is ͮʹͮ, then the extraboard would have ͱͭ slots for picking. Is that the correct number? Without an evaluation of agency needs or how they may have changed over time, it is not possible to say. A healthy balance in terms of work distribution on the extraboard is the goal. In operational terms, a healthy balance would mean minimizing both make-up time and overtime. This indicates that the dispatchers are careful not to assign a piece at overtime that could be worked by a show-up operator. How would an extraboard evaluation be conducted? Each operator on the extraboard is guar- anteed Ͱͬ hours for their work week. Garage dispatchers are responsible for making certain that those Ͱͬ hours are productive and do not have an unusually high amount of make-up time. If make-up time is a regular occurrence, it is a sign there is not enough work for all of the members of the board and that the number of slots should be reduced at the next sign-up. However, if dispatchers are giving make-up time to some while giving signifi cant overtime to other operators on the board, it may be a sign that the board is not being assigned properly— some work assignments are being given to persons who already have enough hours. If there is no make-up time and overtime for the average extraboard operator is becoming excessive, then the extraboard probably needs to be expanded by one or more operators at the next sign- up. relief run A run that is available as a result of other operators’ day off selec- tions. Some rosters are made up of several different relief runs.

Chapter 6. Rostering 6-49 Level ͯ. Advanced Rostering An obvious benefi t of including part-time operators on the extraboard is that make-up time is not necessary, providing garage dispatchers with somewhat greater fl exibility and possibly re- ducing agency cost. However, not all agencies that use part-time operators allow them to be included on the extraboard. There are great diff erences in contract provisions related to both the extraboard and the use of part-timers, making it diffi cult to provide useful information that would apply broadly. One factor that needs to be considered in an extraboard evaluation is that extraboard opera- tors are additional “bodies” that receive the benefi ts entitled to their position—benefi ts that come as a cost. The cost for having extra “bodies” must be weighed against the cost to the agency and the “psychic cost” of excessive overtime and stress on your regular assigned opera- tor staff . Of course, there are legitimate reasons for extraboard operators to gain overtime. A principal reason is working regular runs which have overtime embodied in them. A healthy rotating board will have a mix of regular open runs and Show-up assignments. Some agencies have rules that an operator who did not get work yesterday is always “fi rst out” for any run that needs to be fi lled. This is their way of assuring that the work assigned gets balanced to the greatest extent. A typical approach to sizing the extraboard is to look at the needs of each garage and fi gure what the average number of extraboard operators should be. You could start with the following list, which is intended to cover weekday work: Average number of daily absencesͭ ͭͳ Vacation reliefsͮ ͱ Open runsͯ Ͳ MiscellaneousͰ Ͱ Number of balanced AM and PM trippersͱ ͭ͵ Total ͱͭ Notes: ͭ This is the number which experience over the year or the particular season of the sign-up tells us is needed for a typi- cal weekday. ͮ This is based on the number of vacation slots that were made available during the week in question or an average for all of the weeks of the sign-up if the number varies by week. Some systems provide a steady number of vacation slots during every week in the year, while others vary the amount, generally making many more weeks available during the summer months, when most operators want to take vacation, than in the winter. Those systems typically cut both peak line service and school trippers during the summer and can aff ord to let more operators off .

Level ͯ. Advanced RosteringChapter 6. Rostering 6-50 ͯ Open runs are the known runs which were not bid at the previous sign-up plus the average number of runs that usu- ally get vacated during a sign-up due to retirements, terminations, long-term illnesses and such. Typically, opera- tors may bid on this open work based on their seniority. Other operators may then bid on their work until the least popular work remains uncovered and must be fi lled daily until the next sign-up. Ͱ Miscellaneous is a catch-all for missing operators. It could be the average number of operators who are off on union business or must be covered due to random drug testing or retraining. It can also cover the number of individual days that are made available for personal days, birthdays, or single days of vacation. Also, where the use of FMLA (the Family and Medical Leave Act of ͭ͵͵ͯ) is signifi cant, the average number of operators on this leave can be placed here. Wherever these numbers are known, they can be slotted in here. ͱ A number of agencies fi ll many or most of their trippers at overtime, and this would not apply. However, in balanc- ing system costs between overtime and the number of operators to hire (see discussion following), a certain number of trippers may be desirable to fi ll from the extraboard instead of being bid, and that number is slotted here. While our solution works out to the ͱͭ operators we calculated earlier, the list could just as well have produced an altogether diff erent number. The key is to evaluate the use of the extraboard regularly, so a realistic number is always used in sizing the board.ͭ Full Staffi ng The methodology for calculating extraboard needs is based on an assumption that there are enough trained and available drivers to cover every scheduled assignment and extraboard slot. If not, then dispatchers are constantly in crisis management, mostly resorting to calling in people on their day(s) off to work for missing colleagues. Under this scenario any prior planning of the extraboard is largely thrown out the window, and dispatchers must become diplomatic ambassadors attempting to “sell” the merits of particular assignments to get operators in. Of course, there is always a group of operators who enjoy the overtime and would normally rather work than be off . Almost consistently there is extra work available for this group, but it should not come at the expense of properly sizing and staffi ng the extraboard. Ongoing Analysis Steps To summarize the steps for proper extraboard management: Regularly analyze the number of operator slots assigned to the extraboard. Research and develop numbers to populate a list of absences similar to the one shown on the previous page. Figure out your operator shortages by garage, and make your recruitment plans accord- ingly. 1 The National Center for Transit Research has developed a tool to assist small and mid-sized transit agencies in managing the extra- board. See : http://www.nctr.usf.edu/pdf/77707.pdf. • • • Letting the operator ranks fall below the number needed can be tremendously expensive for an agency and should be cause for major management concern. Tip

Chapter 6. Rostering 6-51 Level ͯ. Advanced Rostering Regularly review the daily extraboard assignments and the weekly payroll fi gures. You are looking to see how well the assignments balanced make-up time and overtime. Repeat this process for all of your agency’s garages, and look for any manpower imbal- ance between facilities. These can normally be addressed easily by moving the junior operators from the garage that has more operators than needed to the one that is un- derstaff ed. During the few months that new operators are considered “probationary,” most agencies are free to move them to any facility where they are needed, with the proviso that additional training on some diff erent vehicle types may be required. Example of Assigning Extraboard To demonstrate how extraboard rotation and assignment work, we present a practical exam- ple. Following is a list of ͮͬ operators assigned to the extraboard. They are arranged in order of their seniority, and the day shown is theoretically the fi rst weekday of a new sign-up; say, Monday, September ͭͳ. During the period when the sign-up is in eff ect, some operators will retire, some who are currently on short-term or long-term leave will return to work and be slot- ted onto the board in seniority order, and some of the positions left vacant will not be fi lled if there is a shortage of operators. Position Name Seniority Date ͭ Fillmore, M. ͭͮ/ͮʹ/ͳͲ ͮ Adams, J ͮ/ͭͯ/ͳʹ ͯ Harding, W ʹ/ͭͮ/ʹͬ Ͱ Nixon, R ͭ/ͭͰ/ʹͭ ͱ Jackson, A ͵/ͭͬ/ʹͰ Ͳ Cleveland, G ͯ/ͭͬ/ʹͲ ͳ Arthur, C ͵/ͬͲ/͵ͬ ʹ Monroe, J ʹ/ͭͱ/͵ͮ ͵ Taylor, Z ͱ/ͬͯ/͵Ͱ ͭͬ Hayes, R. ͭͭ/ͬͱ/͵ͳ ͭͭ Garfi eld, J Ͱ/ͭͳ/͵ʹ ͭͮ Hoover, H ͭ/ͬͯ/ͬͬ ͭͯ Coolidge, C ͮ/ͮͮ/ͬͭ ͭͰ Johnson, L Ͳ/ͭͯ/ͬͮ ͭͱ Grant, U ͮ/ͮͰ/ͬͰ ͭͲ Ford, G ͭͬ/ͬͭ/ͬͲ ͭͳ Harrison, W ͱ/ͭͱ/ͬͳ ͭʹ Adams, JQ ʹ/ͭͰ/ͬͳ ͭ͵ Jeff erson, T ͭͮ/ͭͬ/ͬͳ ͮͬ Van Buren, M ͭͮ/ͭͬ/ͬͳ The other piece of information we need is a list of open work to which the extraboard opera- tors will be matched. On this particular day, the list is made up of the following: • •

Level ͯ. Advanced RosteringChapter 6. Rostering 6-52 Work Assignments for Tuesday, September ͭʹ Report Time ͭ Show-up Ͱ:ͬͬ AM ͮ Show-up Ͱ:ͭͱ AM ͯ Show-up ͱ:ͬͬ AM Ͱ Run ͮͭͮ ͱ:ͭͮ AM ͱ Vacation Hold-down ͱͬͱ ͱ:ͮͮ AM Ͳ Show-up + Run ͯͲͬ͵ ͱ:ͯͬ AM ͳ Run Ͳʹͬͭ ͱ:ͰͲ AM ʹ Run ͭͱͭͬ ͱ:ͱʹ AM ͵ Show-up + Tripper ͵ͱ͵ Ͳ:ͬͬ AM ͭͬ Vacation Hold-down ͮͲͭͰ Ͳ:ͭͬ AM ͭͭ Vacation Hold-down Ͳͭͭ Ͳ:ͮͳ AM ͭͮ Run ͭͱͭͳ ͵:ͮͮ AM ͭͯ Run Ͳͭͮ ͭͮ:ͭͬ PM ͭͰ Show-up + Tripper ͭͲͮͮ ͭ:ͯͬ PM ͭͱ Run ͯͮͬʹ ͭ:ͯͲ PM ͭͲ Show-up ͮ:ͬͬ PM ͭͳ Run ͭͬͭͰ ͯ:ͯͮ PM ͭʹ Run ͰͰͭ͵ Ͱ:ͮͳ PM The fi rst thing you notice is there are ͮͬ operators on the extraboard and only ͭʹ pieces of work to assign. That is a positive position to be in and is not the normal case, to be sure. But we can use it to illustrate principles better than the “usual” case where there is much more work to be fi lled than there are operators. The work falls into broad categories. Experience has indicated we need at least fi ve Show- up positions with show-up times close to where they will be slotted. This gives us coverage throughout the AM pull-out period. We also know that not all Show-ups will be given work, so we can safely assign an open run and a tripper to two of the Show-ups. These are scheduled to pull out about an hour after the on duty time and, if additional people call in sick or miss out, there is suffi cient opportunity to reassign them to another run and get someone else in to work the previously assigned run. miss out The term applied when a scheduled operator does not report on time for his/her assignment.

Chapter 6. Rostering 6-53 Level ͯ. Advanced Rostering Another category is Vacation Hold-downs. This garage apparently will allow three operators to be off for vacation each week, so these assignments are being covered by the board. Some agencies separate vacation coverage from the rest of the extraboard, but here it is just one more requirement for the board to hold down. The runs shown are ones that either did not get bid at the time of the last sign-up and have been left for the dispatcher to fi ll on a daily basis or have operators assigned who have called in to get on the Sick List. In the latter case, the dispatcher will have to fi ll the run until the opera- tor calls again and advises that he/she will be returning to work. Some categories of work are not listed. One or more operators may be assigned for the day for retraining following an accident, or may be listed for a court appearance, or may have asked off on union business. The dispatcher typically will add these in as the board assignment pro- gresses. Doing the Assignment Remember we said that the operator line-up on the board was for the fi rst day of the new sign-up on a mythical Monday, September ͭͳ. The garage dispatcher on that date will prepare the board for assignment for Tuesday, September ͭʹ. The fi rst duty is to rotate the board. This agency rotates one position per day. So, M. Fillmore drops from the number one position down to the bottom at ͮͬ. All other operators move up. The top operator is now J. Adams. The dispatcher then aligns the work in time out order (just as we displayed it above). He/she makes a last minute check of the sick list to see if there are any additions or deletions to it (typically operators have a deadline to get on or off the sick list which is about an hour before the next day’s assignments are posted). Any changes would be refl ected in the mix of runs to fi ll and the time order of the work to be assigned. Below is our fi nished extraboard, which includes the match-up of the open work and also some added work for individual operators as mentioned above. Position Name Assignment Report Time ͮ Adams, J Show-up Ͱ:ͬͬ AM ͯ Harding, W Show-up Ͱ:ͭͱ AM Ͱ Nixon, R Show-up ͱ:ͬͬ AM ͱ Jackson, A Run ͮͭͮ ͱ:ͭͮ AM Ͳ Cleveland, G Vacation Hold-down ͱͬͱ ͱ:ͮͮ AM ͳ Arthur, C Show-up + Run ͯͲͬ͵ ͱ:ͯͬ AM

Level ͯ. Advanced RosteringChapter 6. Rostering 6-54 Position Name Assignment Report Time ʹ Monroe, J Run Ͳʹͬͭ ͱ:ͰͲ AM ͵ Taylor, Z Run ͭͱͭͬ ͱ:ͱʹ AM ͭͬ Hayes, R. Show-up + Tripper ͵ͱ͵ Ͳ:ͬͬ AM ͭͭ Garfi eld, J Vacation Hold-down ͮͲͭͰ Ͳ:ͭͬ AM ͭͮ Hoover, H Vacation Hold-down Ͳͭͭ Ͳ:ͮͳ AM ͭͯ Coolidge, C Run ͭͱͭͳ ͵:ͮͮ AM ͭͰ Johnson, L Show-up ͭͭ:ͬͬ AM ͭͱ Grant, U Run Ͳͭͮ ͭͮ:ͭͬ PM ͭͲ Ford, G Show-up + Tripper ͭͲͮͮ ͭ:ͯͬ PM ͭͳ Harrison, W Retraining—Report ʹa-ͱp ͭʹ Adams, JQ Run ͯͮͬʹ ͭ:ͯͲ PM ͭ͵ Jeff erson, T Show-up ͮ:ͬͬ PM ͮͬ Van Buren, M Run ͭͬͭͰ ͯ:ͯͮ PM ͭ Fillmore, M. Run ͰͰͭ͵ Ͱ:ͮͳ PM Note that one operator, W. Harrison, was assigned to retraining leaving only one operator with no assignment. L. Johnson was subsequently assigned an ͭͭ:ͬͬ AM Show-up, which is a nice- to-have protection slot for any afternoon work that might miss out. We would expect this to be a luxury not usually available to the dispatcher. In such cases, he/she would give an ͭͭ:ͬͬ Show-up to one of the earlier extraboard operators who might not have received any work. Why L. Johnson? Because he was the next operator in seniority that would match up with the ͭͭ:ͬͬ reporting time. Had the time been made later, an operator who was lower on the board would have caught the assignment. This is the simple look at assigning extraboards. A number of agencies have various rules gov- erning how the work is given out that are more complicated than shown here. One example rule is that any operator(s) who do not receive a day’s work (a run or equivalent) today are automatically guaranteed it for tomorrow (or their next working day). In this case, operators will get out of strict seniority order for this and similar instances. The goal is to spread the work around and minimize the giveaway of unproductive make-up time to get each operator to his/ her Ͱͬ hour weekly guarantee.

Chapter 6. Rostering 6-55 Level ͯ. Advanced Rostering Holidays Holidays can be a big issue when preparing the runcut for the pick. The labor agreement or past practice (or both) dictates how holidays are addressed. Most agency practices regarding holidays fall into one of the following three options: Operators are required to work the holiday if their run operates on the holiday Holiday work is posted and picked separately The agency determines an appropriate level of service (e.g., Saturday or Sunday) for the holiday, and all operators who have picked that day are required to work on the holiday More complex situations have evolved from these basic options. One agency had so many op- erational problems running Saturday schedules on certain holidays that it decided to operate weekday service (modifi ed by pulling a few runs) and pay holiday premium. Another agency assigns all employees two letters from A to G (corresponding to Sunday through Saturday) at the time of initial hire and requires them to work a holiday if the day matches the employee’s letter. Operators need to know the holiday consequences when they pick in order to avoid grievances later. At many agencies, these practices are so traditional they are not well documented, but schedulers should be aware lest their actions have unintended eff ects. The scheduler should also be aware of the range of options to aid in the assessment of whether the agency’s current practice can be improved. • • •

Level ͯ. Advanced RosteringChapter 6. Rostering 6-56 LEVEL 3 End of Rostering. Rail Scheduling begins on the next page.

Next: Chapter 7 - Rail Scheduling »
Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s Transit Cooperative Research Program (TCRP) Report 135: Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling explores information on available scheduling tools and techniques and their capabilities. Also, the report provides guidance to transit agencies on a variety of scheduling issues typically faced in a transit operating environment.

TCRP Report 135 is an update to TCRP Report 30: Transit Scheduling: Basic and Advanced Manuals and addresses contemporary issues in transit scheduling. The appendixes to TCRP Report 135 were published as TCRP Web-Only Document 45: Appendixes to TCRP Report 135: Controlling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling.

An interactive scheduling manual is available as an ISO image. Links to the ISO image and instructions for burning a CD-ROM from an ISO image are provided below. Once a CD-ROM has been made with the ISO image, open the folder on the CD-ROM called Interactive Scheduling Manual and click on Transit_Scheduling_Lessons.pps.

Help on Burning an .ISO CD-ROM Image

Download the .ISO CD-ROM Image

(Warning: This is a large and may take some time to download using a high-speed connection.)

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!