In 2020, the COVID-19 crisis led many companies to fundamentally rethink their ways of working. At Thales Digital Factory, we switched from an organization based on a 100% onsite working to a 100% remote working during the quarantine. When health measures eased a bit, a hybrid working system came in. An employee has now the opportunity to work totally from home or to come to the office once a week if needed. It is an attractive balance in many ways: better resilience of the company in the event of a crisis (epidemic, transportation strikes, etc.), better development of employees (reduction of commuting time) and need reduction in office real estate.
While testing this system, we realized that this flexibility brought up other issues:
Thales Digital Factory is located at WeWork’s coworking space in Paris, where companies can rent areas in the building to establish their office. This solution is one of the great ways to satisfy our need for flexibility as we can either rent more office space in case of an employee volume increase or return offices when not needed. When the quarantine was confirmed over time, we made the choice to return an entire floor of the space we occupied, representing about 50% of our total workspace.
This obviously allowed us to make operational savings! But organizing a rotation of Thales Digital Factory teams in a space that has become very limited following the closure of a floor and the distance measures (1.5 meters between each employee) areis not an easy thing to run. Indeed, we had to consider several operational constraints like health measures, team preferences and individual need. In the beginning, it took about a day for our Human Resources team to manually work out a planning. In these conditions, it was not easy to have a dynamic schedule that could be updated quickly. Each modification (depending on daily hazards or in adaptation to new health measures) could engage a member of the HR team for several hours.
Hence, we decided to provide a solution to this problem; an algorithm able to automatically generate the Factory’s schedule in an optimized way and by ensuring that it is as adaptable as possible.
In order to simplify the process, we used standard optimization and operations research tools (generic optimization solvers) as the building blocks of the algorithm. With these tools, we were able to focus on the core of the problem and to model the problem expressed by the user. In a few weeks, and after few modifications with HRs, we have been able to provide the first automatically generated schedule, as well as a user interface allowing the constraints and objectives to be simply specified by the end user..
A few months later, from the same basic building blocks, the model evolved to improved computing time (15 min --> 1 min) and new features like the re-optimization of schedules.
Today, even if we see the end of the health crisis, we will capitalize on the agility developed during this period. Indeed, this crisis has made possible to accelerate the thinking on Smart Working and we would like to develop the algorithm in order to respond to these new issues within the Thales Group.
For more information, contact Data Studio: email@example.com