I. Introduction
Crown Melbourne is one of the leading integrated resorts in Australia, consisting of a casino, hotels, restaurants, shopping and entertainment venues. The company reported a revenue of $1,994.8 million for 2017, which decreased 13.7% year-on-year as compared to 2016; similarly, their 2017 net profit of $400.2 million was down 16.5% against 2016 (MarketLine, 2018). Crown Melbourne has identified operations as an area for improvement and have so far implemented programs to boost efficiency and productivity, with the ultimate goal of improving operational performance (Crown Resorts, 2017). As the company employs over 10,000 people, making it the “largest single-site private sector employer” in Victoria (Crown Resorts, 2017, p.15), an effective way for them to reduce costs would be to optimise staff scheduling.
We have examined the staff scheduling problem in the context of one of the hotels at Crown Melbourne – the Crown Metropol Melbourne. The hotel has 658 guest rooms with 92.2% occupancy rate (Crown Resorts, 2017), which shows high demand resulting in large guest turnover. Considering the elevated expectations from a luxury hotel and different requirements across departments, optimal staff scheduling presents an opportunity for significant cost reduction at Crown Metropol. In fact, human resource management was identified as the most pressing issue by hotel managers across various countries, according to a survey (Enz, 2009).
Specifically, Crown Metropol’s objective is to minimise operational costs by optimising staff scheduling across its various departments, while fulfilling the specific and varying needs of each. Mathematical models such as integer programming, metaheuristics or hybrid models combining the two approaches can be used to solve the problem (Rocha, Oliveira, & Carravilla, 2012). However, the nature of the hospitality industry gives rise to a number of complex constraints, including demand fluctuation, multi-skilled employees and length of shift, which will be used as the main criteria to evaluate the journal articles. A detailed list of constraints considered can be found in the table below.
Table 1: Constraints in the hospitality industry
Constraints Description
Number of employees per department Number of people required by each department in order to carry out their operations seamlessly. Some departments like housekeeping require more employees than others like finance or human resources.
Days off A day’s holiday from work, which would have been a normal working day. This also includes all the government and public holidays and non-working days (paid leave, annual leave etc.).
Working hours Employees’ normal or regular hours of work, which can either be fixed or roster based. Full-time employees work an average of 38 hours per week while part-time employees in the hospitality industry work at least eight hours and less than 38 hours per week (“Part-time employees,” n.d.).
Breaks between shifts Employees should have a break of at least 10 hours between the end of one shift/day and the start of the next (“Breaks,” n.d.). This is to eliminate employee fatigue and distress.
On-shift breaks They include rest and meal breaks during working hours. Rest breaks are paid breaks lasting 20 minutes that are counted as time worked, while meal breaks are uninterrupted, unpaid 30-minute breaks that are not considered time worked (“Breaks,” n.d.).
Contract type Hotels employ both full-time and part-time employees depending on job roles. Roles that are client-facing like guest relations engage full-time employees, while roles that do not require specialised skills like laundry or back-of-house operations are staffed by part time employees.
Work experience The level of prior experience of qualifications a person has for a particular position. High-visibility roles such as guest relations and concierge services, executive positions and crisis response teams require highly qualified employees. In addition, during peak business hours or holiday seasons when demand is high, more experienced staff are preferred.
Type of skilled staff An employee can either be single-skilled or cross-skilled. Cross-skilled workers have training in two or more roles to handle different responsibilities as needed. Cross-skill training is usually done between the housekeeping and kitchen teams, and front office and food & beverage teams. This is to make operations flexible, especially during times of peak demand.
Nature of shift Depending on the demands of various departments in a hotel, employee shifts can either be a fixed (human resources, finance, sales) or 24-hour rotating shifts (kitchen, housekeeping, reception/concierge).
Demand fluctuation It is the market interest variation over time as a result of change in consumer demand. In times of high seasonal demand like Christmas or summer holidays, the required number of employees at hotels also increases. In contrast, during off-peak seasons like winter, low footfall reduces the need for manpower at hotels.
II. Methodology
Although staff scheduling is a common human resource problem, there is a lack of quantitative research addressing the problem in the hospitality sector, in contrast to other sectors such as healthcare and transportation that have been extensively studied (Rocha et al., 2012). In order to help solve the problem, we examined ten journal articles from the past ten years that used various approaches and mathematical models to address staff scheduling. Keywords such as ‘staff scheduling’, ‘rostering problem’, ‘tour scheduling’, ‘hospitality’, ‘optimisation’, ‘mathematical modelling’ and ‘integer programming’ were used to search for the articles in databases including ScienceDirect, Business Source Complete and Science Citation Index.
Information about Crown Metropol Melbourne was gathered from Crown Resorts’ annual report, company reports from MarketLine and IBISWorld, as well as the hotel’s official website. Guidelines for the hospitality industry in Australia and regulations on work times and schedules were taken from government sources such as the Fair Work Ombudsman (https://www.fairwork.gov.au).
III. Literature Review
Each of the journal articles was analysed with respect to our problem of optimising staff scheduling to minimise costs. Table 2 provides a brief overview of each article and model used, before a detailed discussion of their strengths and weaknesses.
Table 2: General overview of articles
Journal Article Description
A dissimilarities balance model for a multi-skilled multi-location food safety inspector scheduling problem (Chen & Kuo, 2016) Chen and Kuo (2016) attempt to optimise staff scheduling at a government food safety centre by creating a fair work schedule for employees. The different efficiency levels of various employees and their work preferences are considered, and a two-phase approach is adopted. In the first phase, work shifts are scheduled with employee choice and unbiased scheduling as the key criteria. Based on the results from phase 1, scheduling is done to match employee skills with the task to be carried out in phase 2.
A shift scheduling model introducing non-regular employees for hotel restaurants (Fujita, Murakami, & Amasaka, 2016) Fujita et al. (2016) suggest detailed shift scheduling models to deal with the challenges faced by the service industry during scheduling. These challenges include reducing employment costs and putting non-regular employees in place to cope with fluctuating demand. While most prior shift scheduling models only considered full-time and part-time employees, Fujita et al. (2016), also included non-regular employees, who are further divided into monthly staff and one-time staff. Model 1 only considers monthly staff in addition to full-time and part-time, while model 2 includes an additional layer of one-time staff as well.
Analysis of three mathematical models of the staff rostering problem (Naudin, Chan, Hiroux, Zemmouri, & Weil, 2012) Naudin et al. (2012) aim to analyse the staff rostering problem by comparing three different Multi Integer Linear Programming (MILP) models they developed by using branch-and-price methods (models 1, 2 and 3). These three models give their own set of solutions considering their respective variables and constraints. Researchers conclude by discussing the best suited models for various defined problems and why model 3 is usually preferred.
Cyclic staff scheduling: Optimization models for some real-life problems
(Rocha, Oliveira, & Carravilla, 2013) Rocha et al. (2013) present a general mixed integer programming model to tackle staff scheduling problems that can be adapted and applied to a variety of everyday rostering problems. Their approach focuses on new formulations of sequence constraints and balancing workloads using cyclic staff scheduling. The designed model is then applied to two real-life case studies: a glass manufacturing plant and a continuous care unit.
Integer programming approach based on pattern for a class of staff scheduling problems
(Ohara & Tamaki, 2014) Ohara and Tamaki (2014) design an approach to generate patterns for allocating workers and then propose a model based on these patterns using the column generation approach. The model presents an effective and practical schedule which aims to find solutions quickly. This optimisation model is then applied to a convenience store scenario to determine its effectiveness.
Mathematical models and solution approach for cross-training staff scheduling at call centers (Taskiran & Zhang, 2016) Taskiran and Zhang (2016) have developed an integer programming model for staff scheduling of a workforce in a call centre with a large mix of cross- trained employees. Constraints considered include schedule of shift, breaks and days off. The modelling has been carried out through a two-phase approach; the most favourable workforce mix is determined in the first phase which is then used to model the staff scheduling in phase 2.
Personnel scheduling using an integer programming model- an application at Avanti Blue-Nile Hotels (Kassa & Tizazu, 2013) Kassa and Tizazu (2013) propose an integer programming model to optimize the weekly shifts of engineers in a five-star hotel, so as to ensure efficient usage of staff. Real-world constraints like breaks between shifts, weekly rest for employees and the number of employees necessary for each shift have been incorporated in the modelling process.
Scheduling part-time and mixed-skilled workers to maximize employee satisfaction (Akbari, Zandieh, & Dorri, 2013) Akbari et al. (2013) formulate a staff scheduling model for part-time and cross-trained workers, with varying productivity levels of workers in a day due to fatigue considered as the key criteria. Factors such as worker’s choices, their experience levels and their availability are used as constraints when framing the model, with the goal of maximising employee satisfaction. An integer programming model along with heuristics like simulated annealing (SA) and variable neighbourhood search (VNS) are used to find solutions near the optimum.
Scheduling restaurant workers to minimize labor cost and meet service standards (Choi, Hwang, & Park, 2009) Choi et al. (2009) try to find the right balance between overstaffing and understaffing by creating an integer programming model for staff scheduling in a Korean restaurant. A ratio of 6:4 is assumed as the ideal mix between full-time and part-time employees when solving the problem, while basic constraints such as fulfilling daily minimum staffing requirements are considered. The model-generated results are subsequently compared to the existing schedule, verifying that the IP model helped reduce employee costs while maintaining service standards.
Simulation and analysis of staff scheduling in hospitality management (Kadry, Bagdasaryan, & Kadhum, 2017) Kadry et al. (2017) create a simulation model using Arena to optimise staff scheduling in a hotel during high-demand season without hiring additional employees. It aims to reduce the time visitors wait in queues by using the model to deliver effective customer service via optimal scheduling. The study mainly focuses on the reception and housekeeping departments during high season.
IV. Strengths and Weaknesses
The strengths and weaknesses of the journal articles have been analysed generally in relation to the constraints outlined earlier. For an in-depth analysis of the pros and cons of each article, please refer to Appendix 1.
Strengths
The journal articles highlight the importance of scheduling and human resource management pertaining to the hospitality industry, with staff and tour scheduling problems the most commonly discussed.
One of the most important things to consider when solving any scheduling problem in hospitality is providing round-the-clock service. This has been addressed by Chen and Kuo (2016) and Rocha et al. (2013) who study the problem in a continuous 24-hour environment. Kassa and Tizazu (2013), on the other hand, modelled different shifts with equal duration in each day of the week. They considered the limit workload and shift sequence, assuming at least one employee to be available at all times, while building a model which was aimed at maximising the work done (Kassa &Tizazu, 2013). Rostering employees throughout the day on a 24-hour schedule is particularly important to a luxury hotel like Crown Metropol Melbourne, where customers demand a high level of service at all times.
Employee skill levels and assigning the right task to the right person is another important factor for the hotel. Ohara and Tamaki (2014) emphasise the importance of the skill-mapping constraint, i.e., allocation of a job to only staff members who have the required skill. Taskiran and Zhang (2016) address the common hospitality scenario of hiring a cross-trained workforce with flexible schedules to adapt during seasonal demand variations and propose a model that incorporates assigning cross-trained staff when shift scheduling.
Naudin et al. (2012) address everyday constraints such as task allocation to every employee, maximum daily work hours, minimum daily rest period, weekly working hours and days off in the model to find a solution to the scheduling problem. These real-life constraints are extremely relevant to Crown Metropol to help them optimise their staff schedule.
[NOT YET FINISHED]
References
https://www.fairwork.gov.au/employee-entitlements/types-of-employees/casual-part-time-and-full-time/part-time-employees#2073-2079
https://www.fairwork.gov.au/employee-entitlements/hours-of-work-breaks-and-rosters/breaks#2192-2198