For manned aircraft, loss of control in flight and controlled flight into terrain are two main causes for aviation accidents. Recently, the authors have developed systems that autonomously execute recovery strategies to rapidly restore nominal flight without overcommanding the aircraft, exacerbating the upset condition, or endangering nearby structures and vehicles. The opportunity exists to extend the automated recovery system (ARS) by incorporating it in manned aircraft in a way that collaborates with the crew about corrective actions and utilizes crew expertise in real-time to provide the most effective recovery. The proposed research will develop and demonstrate such crew-specific extensions to show the benefits of a collaborative human/automated (H/A) upset recovery system. Metrics will be defined to evaluate mixed H/A team performance both in terms of performance and crew experience, and a variety of levels of autonomy will be implemented including manned recovery, ARS advisory mode, human oversight of ARS autonomy, and fully autonomous mode. A key goal is to demonstrate that a mixed H/A mode will provide significant advantages over what can be achieved by either the pilot or the fully autonomous system alone. Toward this end, human-in-the-loop hardware demonstrations will be used to demonstrate the benefits of joint H/A approaches and to set the stage for Phase II flight demonstrations.