Accurate solar sail control in space requires persistent control attention. Hence, the need for adaptive, autonomous control. However, control of the thin film sail material is still an enigma, and will continue to be until solar sails have flown in space. As a result, the proposed flight controller for Solar Blade employs machine learning and computer vision algorithms to learn appropriate parameters for navigation and control while in space. By combining high fidelity off-line modeling and on-line learning techniques, the guidance system will achieve precise control of the large sail structure. The learning controllers will in turn be guided by autonomous mission planners which combine high level mission goals, such as desired trajectory, with other factors, such as dynamic solar pressure and spacecraft/Earth geometry, to enable robust operations.

To navigate through space, Solar Blade utilizes a specialized computer vision system, employing measurements of the apparent positions of the Earth, Sun, and Moon, as well as the apparent size of the Earth or Moon to determine attitude, position and velocity of the vehicle in Earth-Moon space. Wide-angle optics and computer vision algorithms achieve a continuous lock on the relative position of the spacecraft.