As the prospects of chronic mechanical assistance for the failing human heart are fast becoming a reality, and patients are indeed returning home to regain a normal lifestyle, the limitations of this technology upon quality of life are becoming more apparent. To address many of these limitations, investigators are developing next-generation ventricular assist devices. Based on turbopump technology, these new devices offer smaller size, greater efficiency (hence smaller batteries), high reliability, and are more cost effective as compared to their pulsatile predecessors. For all the virtues of these new turbopumps, they bring additional challenges. Arguably the most urgent is the need for added """"""""intelligence."""""""" These relatively ignorant devices are highly dependent on feedback-control to provide normal physiological response. The goal of the Phase- II effort proposed herein is to design a robust controller that may be incorporated into these turbodynamic pump systems for clinical use. The primary end product of this program would be a validated algorithm, in the form of firmware that will be embedded into existing rotary pump controller -- capable of maintaining optimal perfusion of the patient under a variety of hemodynamic demands and disturbances, while avoiding deleterious conditions such as ventricular suction.