By the mid-1990's, the U.S. Geological Survey expects to produce spatial data according to its new data model, Digital Line Graph-Enhanced (DLG-E). This new data model currently defines more than 200 unique feature types that describe the geographic phenomena portrayed on the series of 1:24,000-scale topographic maps. Characteristics of features are encoded as attributes, and linkages between features are expressed as relationships. Ultimately, features are tied to the spatial components that represent their location and (or) shape. Developing the ability to manipulate the features that compose the DLG-E world presents many new challenges in the design of a data production system. Primary among these challenges is controlling the attribution and value of each feature type to ensure consistency in data content. Methods are under development at the U.S. Geological Survey to provide automated control over the DLG-E data production process.