Over the past sixteen years, newspaper PCSs (Production Control Systems) have been developed to support newspaper production, so that newspaper publishers can offer the most up to date information to the greatest possible number of readers in a timely manner while keeping production costs low. Up to now, the chief goal has been to produce a predetermined number of copies of a stylized newspaper (using a set layout and set number of pages) as efficiently and accurately as possible. However, in today's information-oriented society, readers and advertising clients have come to expect low cost newspapers that reflect their various interests and demands. To satisfy these, a PCS should aim at providing a reliable and flexible production system that allows immediate changes in format and circulation volume according to article contents, as well as coordinated production (printing and delivery) between locally distributed printing plants. In this paper, a new newspaper PCS architecture and data-processing and networking techniques are discussed based on an autonomous decentralized system (ADS). The architecture provides a system that allows flexible change of the production schedule, while ensuring reliability, because it has autonomous filtering functions that distinguish between different kinds of data, namely control data and man-machine data. The Asahi Shimbun newspaper company currently uses this highly effective system in its Setagaya printing plant.