Astronauts and robots must collaborate effectively to safely and efficiently explore harsh extraterrestrial environments. This research describes a planner/scheduler system to manage a hybrid rover-astronaut team, focusing on real-time contingency response. Collaborative exploration scenarios studied include observation, sampling, and rescue tasks. A search strategy with branching factor control trades planning time with search space completeness. The planner reacts quickly to anomalies when required but also can build optimal plans that minimize deployment time while maximizing science return. A six-wheeled rover was constructed to perform imaging tasks and carry payloads as needed. The rover is equipped with vision-based navigation, control, and path planning algorithms with obstacle avoidance for navigation in unmapped environments. A human astronaut interacts with its autonomous companions and the planner through a laptop GUI. A series of simulation and hardware-based tests evaluate planner performance and astronaut-rover collaboration scenarios.