Staff scheduling and rostering problems, with application in several application areas, from transportation systems to hospitals, have been widely addressed by researchers. This is not the case of hospitality services, which have been forgotten by the quantitative research literature. The purpose of this paper is to provide some insights on the application of staff scheduling and rostering problems to hospitality management operations, reviewing existing approaches developed in other similar areas, such as nurse rostering or examining adaptable problem models, such as the tour scheduling.
Hospitality is not a recent activity. In the social sense of the concept it dates from ancient times, where many societies had traditions of traveler protection and welcoming. King [
Hospitality and hospitality management have been the scope of many research articles, essentially in the social sciences field, where the discussion has been focused on a common and generically accepted definition and on the development of a framework to be the basis of an independent academic discipline.
Although still being often merged with tourism and leisure sector activities, hospitality services are a growing activity sector in a society where customer’s satisfaction and well-being run the market. They usually include hotels, restaurants and other sort of lodging, food and drinks services providers. Due to the specifications of the kind of service provided, hospitality management has to deal with complex variables and constraints. An unpredictable customer demand, a multiskilled workforce, different staff labour contracts’ demands, employees satisfaction and costs minimization are just some of the conditioning issues that an organization has to deal with in order to achieve a flexible, profitable and quality service provision.
A survey undertaken by Enz [
Staff scheduling and rostering are typical problems to solve within this area. There is however a big lack of published articles focusing these problems applied to the hospitality sector, as realized by Ernst, Jiang, Krishnamoorthy and Sier [
The aim of this paper is to share some findings on the application of staff scheduling and rostering problems to hospitality management operations. An overview on how hospitality management, staff scheduling and rostering topics are discussed in the research literature is given, reviewing existing approaches developed in similar application areas as well as analyzing other problem formulations, with analogous characteristics, that are potentially adaptable to the problems under discussion.
One of the reasons for the lack of research articles focusing staff scheduling problems in hospitality is perhaps the lack of a consensual and generically accepted definition of the activity itself. Etymologically, the word hospitality, in Latin hospitalities, has its origin in hospes or hospitis (genitive), which means foreigner or guest. Dictionary definitions include “cordial and generous reception of or disposition toward guests” [
In the industrial context, the term hospitality has been adopted mainly in the English-speaking countries to refer to the activity of hotels, restaurants and other sort of lodging, food and drinks services’ providers, whether it takes place in a public/commercial or in a private/social context. Lashley [
In a first essay, hospitality management would then be intuitively defined as the management of those hospitality activities. In concordance, Brotherton and Wood [
In the academic community, researchers have been seeking out the development of the specialist discipline of hospitality management that would embody a theoretical framework and link it to the industry sector, but the lack of a consensual definition of hospitality has effectively been a barrier both to research progress ([
King [
Believing that hospitality is a time evolving phenomenon, i.e., that hospitality’ characteristics change over time, Brotherton [
The comprehensive approach of studying the commercial hospitality activity from a wider social sciences perspective has indeed been quite controversial, as it turned out to happen after the publication of the book “In search of hospitality: theoretical perspectives and debates” by Lashley and Morrison [
Slattery [
In his review, Jones [
In a recent article, Ottenbacher, Harrington and Parsa [
Wren [
Several designations can be found in the literature to refer the problem of allocating human resources to work schedules or shifts. Those include staff, workforce, labour, employee or personnel scheduling/ rostering problem. This is a particularly relevant problem for the services activity sector, which is typically labour intensive, and has been the scope of interest of many research groups. A far-reaching review on applications, methods and models of rostering problems may be found in [
In the context of manufacturing organizations, systematic or standard approaches to tackle staff rostering problems are commonly applied. But in the case of service industries it is not so simple. In service operations there is no mass production, every service delivery is unique. While in manufacturing, products have rigid specifications in terms of physical dimensions, shape, etc., services attributes are intangible and service provision is affected by several human subjective factors that distinguish one from another, such as the mood of the receptionist in a hotel in a specific day or the noise of other customers in a restaurant, for example. Another key difference between these two activity sectors is the ability to fit demand. While in a manufacturing company, products can be held in stock allowing a flexible response to demand fluctuations, in pure service operations the “product” is both “produced” and “consumed” at the same time.
The complexity of the staff scheduling and rostering problem in service activities is mainly due to the variability of the demand and also due to its constraints, namely those directly related to employees’ preferences in terms of working or rest days, to incompatibility between staff elements or between tasks, or those related to workload balance or fairness between schedules. These concerns have impact in the employee’ productivity and are particularly relevant in service organizations, where service delivery is directly affected by employees performance, but are usually treated as non mandatory and named soft constraints. Other constraints to this problem, such as legal regulations that are mandatory, are considered hard constraints. Causmaecker and Vanden Berghe ([19,20]) propose a framework for the classification of personnel rostering problems in services. It considers three categories: personnel environment, which includes different types of personnel constraints and skills; work characteristics, which refers to coverage constraints and shift types; and optimization objective. Such a classifi- cation system allows the comparison of problems in terms of hardness and complexity, but also the evaluation of developed approaches.
The rostering problem is often addressed in literature in its several sub-problems, either independently considered or in a combined way. Basically there are three main challenges to tackle within this problem: demand levels that we need to know in order to plan the workforce requirements, a set of activities to organize in shifts and the assignment of the staff elements to shifts and also to the rest days.
Sub-problems of the rostering problem that can be found in literature include demand modeling, shift scheduling, days off scheduling, tour scheduling, crew scheduling, crew rostering, task assignment, shift assignment, staff assignment, rotating or cyclic workforce scheduling, among others.
Demand modeling consists in determining demand levels and translating them into staff requirements for each planning period, for each shift or for each task. It is an important part of the process although it is often tackled at a higher level of more strategic planning decisionmaking than the remaining short-term or operational subproblems of the rostering problem. A generic illustration of the demand modeling phase can be seen in
In some service operations, where customer arrivals are usually random and fluctuate throughout the planning horizon, forecasting, queuing theory and simulation techniques are widely used to determine demand levels and the respective staff requirements. A detailed insight into this topic with application in call centers can be found in [
On the other hand, in activity sectors such as transports, demand is modeled with basis in the requirements of a predefined list of individual tasks to be performed by an employee (driver). Demand modeling in nurse rostering, for example, is based in the number of staff required for each shift, which must be in compliance with predefined service ratios (ex: nurse/patient). In hotels only a part of the demand, corresponding to confirmed
reservations can be known beforehand. The remaining demand determination must be based on historic information and forecasting techniques. Of course, there is also the component of daily check-ins that must be considered.
The problems of crew and tour scheduling arise as the two most explored in literature ([
Crew scheduling and crew rostering are equivalent to the shift and tour scheduling problems respectively, when applied to transportation systems. In these systems the demand is determined based on a set of previously defined tasks. Another difference is the existence of an additional geographical or spatial dimension, usually associated to each task, which in this case can be the trips between two consecutive stops (buses, railways) or flight legs (airlines) that will be combined into round trips or pairings. See [
The tour scheduling problem is typically used in organizations that work continuously, 24 hours a day, 7 days a week, such as hospitals (nurse rostering) or hotels. Schedules are usually defined in a weekly basis.
In situations where the demand patterns repeat in a weekly basis, each line of the work schedule will rotate
through all employees with an offset between them. For example, employee number 2 will have the same line of work in week 2 that employee number 1 had in the week before. This is a cyclic rostering problem or workforce rotating scheduling problem [
Staff scheduling and rostering problems have been explored in literature in many application areas, with emphasis on hospitals and transportation systems [
In an attempt to justify the relevance of staff scheduling for hospitality managers, Thompson proposes a fourstage method for the hospitality industry, which he presents in a four-article publication: forecasting demand [
Choi, Hwang and Park [
The staff rostering problem in hospitality services shares common features with other service activities. It is quite noticeable, for instance, its similarity with the nurse rostering problem, so deeply explored in literature. Considering hospitality in its narrowest sense, as defined in the previous sections of this paper, which includes mainly hotels and restaurants, both problems seek to assign a set of employees to a set of working days, shifts and rest periods in order to satisfy demand levels, taking into account concerns regarding work rules, employees’ skills, availability and preferences. The rostering problem both in hotels and hospitals is typically characterized by an around the clock operation, 7 days per week. The use of different contract types (e.g. part-time) is therefore a common and necessary practice.
Of course the place where it takes place and the set of conditions under which it can be found make the approaches to the rostering problem to be different in a hospital or in a hotel. The nurse rostering problem takes place in a hospital unit—a ward, where usually the different skill categories of the nurse function (e.g. head nurse, regular nurse, caretaker) need to be taken into account. The considered shift types are usually the conventional 7 or 8 hour shifts: early, late and night. Demand is usually determined with basis on desired service levels (e.g. nurse/patient ratios) or/and forecasting techniques and staff levels are defined for each shift and skill category. Workload distribution follows a daily pattern, usually ignoring weekends. Work rules are strict in terms of shift sequence, maximum/minimum number of consecutive assignments for each shift, periodicity of rest days, etc.
In addition to its lodging core operation, which involves several different functions (receptionist, concierge, doorkeeper, cleaning staff, maintenance operator, administrative staff) a hotel usually includes other activities such as restaurant, bar, leisure spaces, etc. The rostering problem in a hotel may also be applied to only a single functional area, for example the scheduling of the reception staff or the cleaning staff. There are situations, however, where staff is multifunctional and so an integrated approach is more appropriate, increasing the complexity of the problem to solve. It requires a high flexibility in terms of shift length, starting and finish times and needs to manage a bigger diversity of employee contract types as well as multiple functions. In what concerns work rules, just like in the case of hospitals, hotels and restaurants are very conditioned by sectorial union agreements or contracts, namely in terms of working and rest periods. In the current globalization context, multinational hospitality organizations, and specially hotels, must follow the motto: “think globally, act locally”, meaning that although having general rules, common to all their units, each unit must have its organization and practice adapted to the local context where they are placed in. Different cultures and different habits usually mean different needs. In the hotels case, the majority of the reclaimed service levels are strategically imposed by its category (number of stars). Staffing needs must be determined based on the historic data, on guest arrival forecasts, on a slack for daily late arrivals but also based on the desired service levels.
In a similar way to the case of nurse rostering, important schedule’s characteristics to take into account in hotels are: coverage, quality, stability, flexibility, fairness and cost [
One adoptable approach to the rostering problem in hospitality organizations, and in hotels in particular, is the tour scheduling problem. Several variations of this problem can be found in literature: models that consider full or/and part-time employees (see for example [
In a 2004 survey, Alfares [
In a survey work, Cheang [
Improvement and constructive heuristics and metaheuristics are alternative approaches to optimization techniques. They tend to model real problems in a more accurate way and to give results in a reasonable time span, although they do not guarantee the achievement of an optimal solution. Goodale and Thompson [
Blöchliger [
In the third of his four-paper work, Thompson [
Although hospitality and hospitality management are subjects that have been quite explored by social science researchers, they are not commonly referred in the operations research literature, or more precisely in the scheduling operations literature. But the truth is that there is a point in exploring rostering and scheduling problems in this area. First of all, because as in any other service activity, the importance of staff expenditure is typically very significant in the total operating costs of a hotel. Secondly, because the quality and efficiency of the service provided by a hotel or a restaurant have direct impact in its customers’ satisfaction, as in few other service activities. The social dimension of hospitality that has been in the hospitality research agenda in the last decades, increases the complexity of staff rostering problems in this activity area. It is no longer only a matter of assuring the required employees’ technical skills, but also of guaranteeing that they have the right personal competences to interact with customers, to understand and satisfy their needs. The staff must be motivated and engaged with the service. Staff scheduling systems shall therefore account for the workforce well-fare, considering employees’ preferences in terms of work and rest days, weekends off and holidays, shifts assignment, shifts change, shifts starting and finishing times flexibility, compatibility or incompatibility with other staff elements, etc. Possible approaches to the staff scheduling and rostering problem in hospitality management, or its subproblems, may be inspired by the work that has been comprehensively done in both tour scheduling and nurse rostering. As exposed before in this paper, nurse rostering and hospitality are two activity areas with many similarities concerning rostering issues. Examples of the few divergences between them include the seasonality, the weekly and daily cycles operation inherent to hospitality activities, which contrast with the typical homogeneous workload distribution of hospitals throughout the year. Concerning solution methods, metaheuristic techniques have been gaining ground on optimizing approaches, although the announced trend is the use of hybrid approaches, taking advantage of the best of more than one single methodology. Thompson [