November 23, 2022

Workforce planning


What if you can automatically generate a work schedule for your employees? That is one of the innovation cases where the Datalab worked on for a couple of weeks. As input data, we had a list of employees and their availability during the week. Additionally, we have an amount of work shifts which need to be filled. With a smart algorithm we try to match this supply and demand. 

The algorithm that we used is a genetic algorithm. A genetic algorithm is a self-learning search and optimization algorithm. The algorithm is based on evolution. We see a solution to the problem as a gene that mutates and reproduces. But, we are not just looking at one gene, but thousands of genes. By generating and reproducing many random solutions, we can use a score function to get a solution that meets all requirements.

In the genetic algorithm of the Workforce planning project, we have to think of the schedule as a gene. We generate thousands of schedules and ensure that they are assessed based on different conditions (e.g. flexible vs permanent staff). The best schedules are chosen and they are reproduced and mutated with each other. This is done until we have a solution that meets all requirements and the score is no longer improved.

The result is a Tableau Dashboard containing the schedule for the week. You can also see per person how many hours they have been allocated and how many contract hours they are on the payroll.

We are looking forward implementing this solution at an OpCo. Are you interested? Please contact any member of the Datalab.

  • Koen Haenen

    Lead Data Scientist
  • Ellen Mik

    Senior Data Scientist
  • Jesper Slik

    Senior AI Specialist
  • Wilte Falkena

    Senior Data Scientist
  • Carlijn Levert

    Data Scientist