The braking system is the main active safety equipment of vehicles. This paper presents a new brake system architecture based on the use of proportional servovalves. The use of servovalves allows for faster and more precise control of the pressure, preventing the wheels from locking and reducing braking distances. A new control block has been developed to determine the optimum pressure in the braking circuit. The new controller is based on the use of fuzzy logic techniques. A Kalman-filter-based estimation algorithm allows for obtaining the adhesion between the wheel and road surface and vehicle speed. These parameters are used in a fuzzy block to detect the road type. Using the road type, an artificial neural network gives an estimation of the slip that produces higher adhesion, and a final block, which is also based on fuzzy logic, determines the optimal pressure in the braking circuit. The performance of the new brake architecture and controller is evaluated by simulating driving on different road surfaces and with real tests on dry and wet asphalt.