Industrial control systems have unique system requirements which often require diverse systems of cooperating components to achieve system goals. System requirements may be specified in the form of tasks which must be executed in sequential order. This paper presents a design perspective in which such problems can be decomposed and implemented by a distributed system in a modular-fashion. A System Scheduler is responsible for all flow control functions. Loosely coupled Task Processors respond to instructions from the scheduler to perform the control algorithm Parallelism, redundancy, and distributed control are presented as methods of improving system performance and reliability.