A rule-based controller for rotary ventricular assist devices was developed to automatically regulate the pumping speed of the device without introducing suction in the ventricle. The control approach is based on a discriminant analysis function that detects the occurrence of suction, providing the input for the rule-based controller. This controller has been tested in simulations showing the ability to autonomously adjust pump flow according to the patient's level of activity, while sustaining adequate perfusion pressures. The performance of the system (suction detector and controller) was tested for several levels of activity and contractility state of the left ventricle, using a lumped parameter model of the circulatory system coupled with a left ventricular assist device. In all cases, the controller kept cardiac output and mean arterial pressure within acceptable physiologic ranges.