Robotic rovers can be used as advance scouts to signifi cantly improve scientifi c and technical return of planetary surface exploration. Robotic scouting, or "robotic recon," involves using a robot to collect ground-level data prior to human fi eld activity. The data collected and knowledge acquired through recon can be used to refi ne traverse planning, reduce operational risk, and increase crew productivity. To understand how robotic recon can benefi t human exploration, we conducted a series of simulated planetary robotic missions at analog sites. These mission simulations were designed to: (1) identify and quantify operational requirements for robotic recon in advance of human activity; (2) identify and quantify ground control and science team requirements for robotic recon; and (3) identify capability, procedure, and training requirements for human explorers to draw maximum benefi t from robotic recon during vehicular traverses and on-foot extravehicular activities (EVA). Our studies indicate that robotic recon can be benefi cial to crew, improving preparation, situational awareness, and productivity in the fi eld. This is particularly true when traverse plans contain signifi cant unknowns that can be resolved by recon, such as target access and station/activity priority. In this paper, we fi rst present the assumptions and major questions related to robotic reconnaissance. We detail our system design, including the confi guration of our recon robot, the ground data system used for operation, ground control organization, and operational time lines. Finally, we describe the design and results from an experiment to assess robotic recon, discuss lessons learned, and identify directions for future work.