NLS will focus on practical feature-oriented work in the field of automatic generalization. The main aim is to increase the use of automatic generalization routines in the different production systems. Secondly NLS will focus on two different scale areas, generalization in medium scale (1:10,000 to 1:50,000) and in small scale (1:250,000 to 1:1,000,000). As a third part NLS also wants to find out in a more accurate way the effects of using automatic generalization routines in terms of saving production costs. In the production line of the database in 1:50,000 scale, the 1:10,000-scale database is the major data source. Today some automatic generalization operators, such as simplification, classification, and collapse, are in use. The aim is to increase the number of operators and to enhance some of the existing routines. The collapse operator has been used for some years now for creating point symbols from area buildings. The existing routines have a large number of limitations such as only buildings with a certain object code are taken care of. Our intention is to enhance this routine so that more can be taken care of automatically. Visualization of geographic information in medium- and small-scale maps leads to spatial conflicts. To try to solve some of these conflicts we are about to develop an automatic displacement routine. It will work on vector data, and the aim is that it will take care of both conflict identification and feature displacement. It will of course have limitations in terms of what kind of object the routine will handle, but our aim is to solve problems step by step instead of solving a major problem at once. The full presentation will focus on the approach to and use of collapse and displacement routines in medium- and small-scale mapmaking.