The project deals with control algorithms directed at optimization of fuel consumption in vehicles from economical/ecological point of view. Bayesian methodology is used. The assumed results of the derived algorithms will be the optimal control variables, namely, pressing the gas and the brake pedal and selection of gear of transmission, based on data continuously measured on a driven vehicle and on external observations. Reducing the fuel consumption is supposed to be reached via tracking the modeled recommended speed used as a set-point. The resulting control system will be implemented inside a test vehicle, where driver will be able to choose or to turn off this system.