Exploration of asteroids and comets can give insight into the origins of the solar system and can be instrumental in planetary defence and in-situ resource utilization (ISRU). Asteroids, due to their low gravity are a challenging target for surface exploration. Current missions envision performing touch-and-go operations over an asteroid surface. In this work, we analyse the feasibility of sending scores of nano-landers, each 1 kg in mass and volume of 1U, or 1000 cm3. These landers would hop, roll and fly over the asteroid surface. The landers would include science instruments such as stereo cameras, hand-lens imagers and spectrometers to characterize rock composition. A network of nano-landers situated on the surface of an asteroid can provide unique and very detailed measurements of a spacecraft impacting onto an asteroid surface. A full-scale, artificial impact experiment onto an asteroid can help characterize its composition and geology and help in the development of asteroid deflection techniques intended for planetary defence. Scores of nano-landers could provide multiple complementary views of the impact, resultant seismic activity and trajectory of the ejecta. The nano-landers can analyse the pristine, unearthed regolith shielded from effects of UV and cosmic rays and that may be millions of years old. Our approach to formulating this mission concepts utilizes automated machine learning techniques in the planning and design of space systems. We use a form of Darwinian selection to select and identify suitable number of nano-landers, the on-board instruments and control system to explore and navigate the asteroid environment. Scenarios are generated in simulation and evaluated against quantifiable mission goals such as area explored on the asteroid and amount of data recorded from the impact event.