Abstract
Recent technological developments try to satisfy the increasing demand of the automotive market for automatic transmission, matching at the same time the conflicting requirements of comfort, performance, fuel economy and cost reduction. An extremely promising system is the robotized gearbox, provided either in semiautomatic or in full automatic ( AUTO) mode. This system is based on servo actuators applied to a standard mechanical gearbox where the gearshifting transient has been optimized w.r.t. driver requests in terms of comfort and performances. By tuning the full automatic mode on the specific and also occasional expectations of the driver, a new concept has been introduced in this automated transmission, the personalization, called the EGO Mode. In this paper such a new concept is outlined describing both the design phase and the implementation phase. In the first one, cluster analysis technique has been used to process data obtained by experimental test on a large sample of drivers, leading to a small number of quantifiable parameters that distinguish a driving style with respect to various environment conditions. In the second one the fuzzy logic has been employed to realize in a simple way these tasks. The robustness of the driving style estimation process is also evaluated.