Combining Optimal and Learning-Based Control Methods for the Manipulation of Spaceborne Objects

Robotics and autonomy continue to push the boundaries of space exploration, but direct interaction with the environment will enable more rich scientific understanding for planetary missions. Additionally, as we look towards future crewed exploration of the Moon and beyond, orbital and surface-based stations will require autonomous upkeep to prepare for and support arriving crew. Robotic manipulation will be integral to supporting these objectives -- allowing robots to interact with the environment and perform rich, dexterous tasks to directly achieve scientific goals and support crew. However, in these extreme environments (such as microgravity and planetary/lunar surfaces), manipulation and mobility control methods need to be highly robust, safe, and function on novel robotic platforms. Therefore, we propose a combination of optimal and learning-based control methods for these manipulation tasks -- combining the computational efficiency and guarantees of optimization methods with data-driven learning techniques to provide robustness against unmodeled disturbances and system dynamics. Through Model Predictive Control (MPC) and Reinforcement Learning (RL), this can be feasible even in cases where evaluating the full behavior of a high-order system is challenging. In particular, we focus on Astrobee (a free-flying robot operating in the ISS) and ReachBot (a cable/boom-driven robot proposed to explore Martian lava tubes), yet we also aim to create generalizable techniques that can apply to future missions and robotic platforms. This will advance NASA's interest in Autonomous Systems and Robotics (ASR) with a focus on Mobility and Manipulation, leading to new capabilities for furthering NASA's science and exploration objectives.

This is early/mid-stage (TRL 2) — the most realistic path in is NASA SBIR/STTR, which funds small businesses and research institutions to develop technology aligned with NASA's needs (equity-free, phased funding). Check whether a current SBIR/STTR solicitation topic overlaps with this project's technology area, or contact the project directly (above) to ask.

None of these are guaranteed paths for this specific project — TechPort itself doesn't have an "apply" button. Reaching out to the contact(s) above with a specific question is usually the fastest way to find out what's actually open.