This paper presents an investigation of flocking by teams of autonomous mobile robots using principles of Swarm Intelligence. First, we present a simple flocking task, and we describe a leaderless distributed flocking algorithm (LD) that is more conducive to implementation on embodied agents than the established algorithms used in computer animation. Next, we use an embodied simulator and reinforcement learning techniques to optimize LD performance under different conditions, showing that this method can be used not only to improve performance but also to gain insight into which algorithm components contribute most to system behavior. Finally, we demonstrate that a group of real robots executing LD with emulated sensors can successfully flock (even in the presence of individual agent failure) and that systematic characterization (and therefore optimization) of real robot flocking performance is achievable.