Energy efficiency represents an important aspect of mechanical design. Despite their long history, gears still play a determinant role in several applications ranging from the automotive, to the aeronautical sectors. The more and more stringent regulations in terms of efficiency have encouraged the gearbox manufacturers to increase the investments to achieve more efficient designs leading to energy saving, reduction of pollutant emission and increased reliability related to the reduction of the operating temperatures. A decrease of the power losses allows also a downsize and a reduction of the weight of the system, with an increase in the power density and performances. Engineering tools allowing a comparison of different design solutions already during the design stage can pave the way to a real transition to a sustainable future. Most available models are based on empirical relations and dimensional analyses resulting to be accurate only as far as the geometry and operating conditions reflect the ones used to calibrate the models. With the developments in computational performances the research started to focus on numerical approaches. However, while most of the numerical approaches have been proved to be sufficiently accurate to capture the power losses of geared systems, the high computational effort required for their application to real gearboxes is still hurting with the industrial practice. Moreover, new phenomena related to new lubricant (e.g aeration, channeling, circulation) could be not captured/simulated with the standard available models. In this paper the latest advancements to overcome both the computational effort issue and the lack of specific models are shown with practical industrial case studies.