Many myths surround the use of optimizers in workforce scheduling, particularly regarding their relevance in self-scheduling and cyclic rosters. To address these questions, we present the three main insights gleaned from our webinars.
1. Collaboratively define what constitutes a ‘good’ schedule
Ask ten people to describe a ‘good’ schedule, and you’ll likely receive ten different responses. To foster a shared understanding, it’s crucial to collaboratively define the key elements of a successful schedule. A schedule must adhere to hard constraints, such as labor laws, qualifications, and contractual obligations. However, balancing soft constraints—like employee preferences and customer needs—poses a challenge, especially when these wishes may conflict. Therefore, alignment is necessary. A ‘good’ roster is not just about satisfying the planner; it also needs to meet the goals of the organization and others. Consider the importance of a ‘healthy roster’ as well.
Once you’ve defined what a ‘good’ schedule looks like, you must be able to evaluate it by measuring KPIs. These metrics provide insights into efficiency and ergonomics and help identify areas for improvement across different departments. This way, you can discuss facts and figures rather than feelings.
"Perhaps the most compelling argument for using optimizers is their ability to create fairer rosters."
2. Optimizers outperform human scheduling
With a clear definition of a good schedule in place (as outlined in tip 1), organizations can begin to compare manual schedules against optimized ones. Experience from various implementations shows that optimized schedules can reduce employee costs by at least 2% by minimizing instances of scheduling over and under the contract hours. This way, more work can be done without extra cost and without working harder.
Moreover, optimizers can better accommodate employee requests while eliminating violations of labor regulations. The key advantage lies in a computer’s ability to evaluate a vast number of scheduling options, much like how computers excel at chess. Additionally, optimizers significantly reduce the time spent on manual scheduling, often cutting planning time by over 50%, thereby streamlining the entire process.
Perhaps the most compelling argument for using optimizers is their ability to create fairer rosters. They distribute workloads evenly, respect all preferences, and ensure equal treatment of employees. With an intuitive app for submitting scheduling requests, employees can easily communicate their needs, reducing ambiguity and enhancing clarity.
3. Relevance of optimizers in cyclic rosters and self-scheduling
Many industries continue to rely on cyclic rosters, where schedules are predetermined months in advance. However, the workload can fluctuate, and tasks may need to be allocated across various workstations with unique qualification requirements. In such cases, optimizers can effectively assign tasks to available employees while adhering to contract hours and individual preferences.
This principle also applies to self-scheduling, where employees create and modify their schedules in the initial rounds. If the schedule remains incomplete after the second round, a planner must intervene to rectify any over- or understaffing issues. While manual adjustments are possible, employing an optimizer can yield a more effective schedule with fewer disruptions.
Moreover, there are further opportunities to leverage math and AI in scheduling. For instance, accurately forecasting workload demands and translating those into necessary shifts with appropriate qualifications is vital. A perfect schedule loses its effectiveness if the required number of shifts is miscalculated, leading to stress or unnecessary costs.
At ORTEC, we’re witnessing a growing trend among our clients to embrace optimizers, whether in conjunction with self-scheduling, workstation scheduling and others. We are proud to highlight the notable successes of clients like Albron (catering) or Maxima MC (healthcare) among others. Given the rapid advancements in AI and mathematical modeling, the capabilities of optimizers are only expected to grow stronger.
