This report reviews ongoing research in modeling and control of assembly systems. A modeling framework is developed for representing assembly tasks as a collection of discrete operations with precedence relations reflecting physical constraints. Devices are mapped into the subtask description leading to an 'operation precedence graph' which is useful for analyzing alternative system configurations and supervisory control structures. Underlying continuous processes determine the tolerance requirements for accomplishing the subtasks. Since reducing uncertainty in the system configuration is the fundamental objective in assembly tasks, two alternatives for representing uncertainty are defined and illustrated by examples. These representations of uncertainty are discussed with respect to alternative feedback control strategies for satisfying tolerance constraints. The concluding section identifies directions for future research.