Electricity accounts for a large proportion of operating costs. Therefore, controlling and monitoring the energy consumed is important for improving the quality of energy use and minimizing losses. This task is most convenient for supervisory systems using wireless sensor networks, which need to operate in an industrial environment, exposed to low-frequency radiation from industrial substations and medium-voltage electrical fields. The aim of this work is to carry out load shedding defined by a genetic algorithm and, using a wireless sensor network and a supervisory system, to optimize production even with limited energy consumption.