Dipartimento di Ingegneria informatica, automatica e gestionale

Nowadays, a key success factor for many large enterprises is the ability of properly managing labor cost and timetables. This is the reason why workforce planning and scheduling tools are now getting more and more developed. Two are the typical issues arising in such applications: the first is related to the medium and long-term goal of estimating the amount of workers that the company will require in future periods. The second, mostly linked to short-term operations, involves the assignment of human resources to activities in order to meet deadlines and industrial plans. In practice, to conduct a complete analysis and evaluate the effectiveness of a solution it is important to take into account both time and financial objectives, considering not only the need of reducing durations and delays but also the ability to do so within reasonable budgets. The result is a trade-off problem looking at the same time at avoiding resource underutilization and incapacity to comply with due dates. In the following, we present a new approach to solve the workforce scheduling problem in complex applicative contexts such as manufacturing and logistics, characterized by the simultaneous processing of several activities, the occupation of wide areas, the coexistence of independent workloads, the use of advanced machineries and, above all, the employment of different types of operators, having various abilities and experience levels. Standard approaches usually address this issue by defining distinct planning, scheduling and allocation problems. However, within the considered context, the problem of providing the right number of workers with the right skills at the right time is inherently linked to the schedule of the activities. For this reason, we rather propose a strategy to tackle all these aspects at the same time, taking into account a reasonable time horizon. As a result we obtain a large problem requiring not only a proper representation of processes complexity, but also a feasible assignment of operators to tasks and an optimized activities scheduling. In what follows, the structure of the problem is formalized and a specialized simulationbased decomposition framework is proposed.