This paper summarizes work on a Kinematic Cellular Automata (KCA) approach to nanoscale replicating robots. Initially funded by NIAC at General Dynamics, the findings include the following: A precise formulation of the essential problem of machine replication can be described as follows: in a well-defined environment, a KCA Indirectly Replicating System (IRS) must assemble basic parts into symmetrical facets of modular dynamic cells configured into a hierarchy of subsystems: Transporter, Connector, and Controller. A practical IRS needs a specific hierarchy that can be broken down in terms of structure, functionality, and control mechanisms. KCA offers a level of indirection that lowers the complexity (to less than that of a Pentium 4). Replication at a molecular level may improve cost and complexity by more than seven magnitudes. The results of modeling different aspects of the KCA IRS are also presented.