Content area
Full text
Introduction
Personnel scheduling problems may be classified as days-off, shift,or tour[1 ]. Days-off scheduling isconcerned with the specification of employee non-work days across aplanning horizon[2 ,3 ,4 ,5 ,6 ,7 ,8 ,9 ,10 ]. Shift scheduling specifies daily startand stop times, and may also include the determination of the durationand placement of rest and/or meal breaks[11 ,12 ,13 ,14 ,15 ]. The labour tour scheduling problemconsists of the specification of both shifts and non-work days foremployees across a planning horizon which is typically one week[16 ,17 ,18 ,19 ,20 ,21 ,22 ,23 ,24 ,25 ,26 ,27 ,28 ,29 ,30 ,31 ,32 ,33 ].
It is likely that the majority of service organizations are actuallyfaced with the labour tour scheduling problem. Historically, the besttour scheduling performances, relative to a goal of minimization of thecost of scheduled labour, have been achieved by methods which have usedmini- and/or mainframe computers[20 ,21 ,23 ,24 ,25 ,26 ,27 ,28 ,29 ,33 ]. Alternatively, microcomputer-based scheduling methods typicallyhave not provided efficient labour tour scheduling solutions. This isespecially unfortunate for service organizations, since they usuallyhave neither the necessary expertise nor the mini- or mainframecomputer equipment to achieve the best solutions[19 ,20 ]. As a result, many service managersspend at least one day per week in the manual development of labourschedules[34 ].
The objective of this research was the development of efficient,microcomputer-based working set procedures for full-time (FT)and mixed-workforce (MW) tour scheduling in a discontinuous (less than24 hours/day) environment. As used in the literature, a MW is one whichencompasses both FT and part-time (PT) employees. Our previousexperience in assisting service operations managers with theirscheduling problems suggested that the new working set methodologyshould not require the service operations manager to specify a workingset size.
Unfortunately, the majority of existing working set methods dorequire the specification of the working set size. This study presentsan alternative approach in which the working set is specified as the setof tours associated with a heuristic solution to the tour-schedulingproblem. Specifically, we propose the micro-computer implementation oftwo-phased heuristic solution methods for the generation of workingsets. For the ET experiment, we utilized the two-stage solutionprocedure developed by Bechtold and Showalter[22 ]. The application of branch-and-bound integer programming(for a maximum of 1,000 integer iterations) to the resulting workingsets yielded integer-optimal solutions for...