The Workforce Scheduling API offers two solvers using the SolveShiftGeneration and SolveShiftScheduling methods. The shift generation solver creates a set of shifts that maximizes coverage of employee demands over a given time horizon. In contrast, the shift scheduling solver assigns specific employees to a set of predetermined shifts. Each solver can be used independently. In particular, the shifts to which employees are assigned in the shift scheduling solver could be specified by the user or could be generated with the shift generation solver.
Shift generation solver
The shift generation solver selects a set of shifts such that the total number of employees required to cover a given demand profile is minimized.
The input of the solver is a request that describes shift templates and employee demands. Shift templates determine the rules that the solver must follow to create shifts. Employee demands describe the number of employees overtime, with a time granularity as short as 1 minute.
The output of the solver is a response with a list of shifts and the corresponding number of required employees per shift. Whenever it is not possible to cover all demands, a list of demand coverage violations is also provided to the user for convenience.
Features
For an in depth explanation of every feature of the solver, refer to the API documentation. The following summarizes the most popular features of the solver.
Shift templates
A shift template primarily determines the duration and potential start times for a shift on any given day. Additionally, it is possible to specify a maximum number of employees per shift template per day and the number of days off per week.
Shift templates can also specify a list of event templates to represent scheduled interruptions such as breaks or employee trainings. Event templates determine the start time (relative to the shift start time) and the length of an event.
Employee demand fluctuations
Demand for employees can fluctuate over time. In particular, employee demands can be specified over time intervals as short as 5 minutes or as long as the planning horizon. Time intervals for which there is not a specified demand are equivalent to having a demand of zero employees.
Shift scheduling solver
The shift scheduling solver optimizes the assignment of employees to shifts so that scheduling constraints are met (or violations thereof are minimized) and employee preferences are taken into consideration.
The input to the solver is a request that describes employees, shifts, coverage requirements, and budget requirements. For each employee, roles, skills and scheduling constraints can be specified to define what a valid schedule per employee looks like. Shifts determine when and where employees can work, while coverage requirements determine the demand for different employee roles (and skills) over several time spans.
The output of the solver is a response with employee-shift-role assignments that summarizes for every employee, the shifts and role (by shift) to which they are assigned.
Features
For an in depth explanation of every feature of the solver, refer to the API documentation. The following summarizes the most popular features of the solver.
Employee scheduling constraints
Specifies rules or limits that the schedule of a particular employee should satisfy. In particular, it supports:
- Minimum / Maximum working time
- Minimum / Maximum consecutive days
- Minimum / Maximum number of shifts
- Minimum rest time between shifts
- Resource constraints
- Shift requests to work (or not) on specific shifts
Employee preferences
Specifies a numeric preference that an employee has for a specific shift, and the sum of preferences across the roster is maximized.
Coverage requirements
Specifies a number of employees that are required to be working over several time windows and for specific roles. Coverage requirements can align with the given list of shifts, but they can also define more fine-grained time windows of the required number of working employees.
Budget requirements
Specifies a contract for the employees with regular and overtime hourly rates, coupled with global budget constraints. A budget can either be a maximum value for workforce costs or an indication that such a cost needs to be minimized.
Constraint priorities
Every employee scheduling constraint, coverage requirement and budget requirement is given a priority level (mandatory, high, medium, or low). All constraints with a mandatory priority must be satisfied by the solver. Constraints with any other priority can be violated by the solver, but these violations are minimized according to the priority order. Employee preferences are considered by the solver after constraint violations in every priority level have been minimized.
How to get started
To get access to the Workforce Scheduling API, follow the trusted testers instructions and then follow the setup guide to obtain an API key. After obtaining access to the API, use the API examples to get started. Should you have questions, contact or-workforce-scheduling+contact@google.com.