Automation of Map Production and Updating Processes - Why it Matters

Over the past several years, there has been a lot of discussion about why automation in cartography is important or whether or not it is even possible to automate important, recurring tasks in the production and updating of maps. The main reasons why a further exploration of automation in cartography makes sense is twofold: one, because it offers a possibility for timely and consistent creation and updating of map and data products in a landscape of rapid change and two, because it opens up the potential for new uses of location-based data and maps.

Automation, in any context, makes sense when it reduces, eliminates or speeds up steps in a process. In the case of cartography, these are the recurring steps in map creation and updating which are done manually or with the aid of computers. These steps include visualization of the geo-data, generalization to make the data readable, placing text and recurrring updating of data.
 


Next to text placement, generalization is the core activity of cartography. Generalization makes the elements on a map readable in spite of a limited amount of space in which to visualize them. Critical elements are accentuated and less important elements are simplified or left out all together. Generalization is an intensely complex process, which requires adherence to a multitude of contextual, geometric and topology rules and was thought to be a process that could not be automated. (Käuferle 2010).

Generalization is done according to constraints that are determined by map scale, the map thematic (also known as map context) and the data density, among others. These contraints are applied to generalization tasks such as displacement, scaling, selection, etc. The constraints are used by the system to identify possible conflicts, carry out the correct solution and quality control. Quality control ensures that the constraints and rules are met and that no further manual intervention is required for those elements which meet the constraints. This is true for both manual and automated generalization. The automatic generalization in axpand mimics the generalization steps and constraints that were traditionally carried out manually by cartographers. This means that the automatic generalization in axpand allows for the creation of new products simply through the adjustment of the constraints. axpand carries out the steps in generalization entirely automatically and in a fraction of the time it takes to do them manually. The results can be repeated and are always consistent for the same constraints. Results show that up to 80% of the objects are correct after generalization. This means that only up to 20% of the objects that do not fulfill the constraints require manual intervention. The automation components in axpand can be integrated seamlessly into all leading GIS infrastructures. 

The following graphic shows maps that have been automatically generalized with axpand.

In a case study and later for use by German authorities, an axpand automation process was constructed to reduce the time for producing and updating a digital topographic map series created with German-specific AAA© data. The goal for time reduction was 80%. This was in spite of limited resources available to work on the map series. This process can be integrated seamlessly into any existing AAA© system infrastructure and business processes. The investment in existing systems remains intact. The investment in the axpand automation can be recovered with time and cost savings after the production of the first map series.
 
The following calculation example from the field shows the cost and time required for manual production of a 25:000 map series.

The positive impact of axpand automation on time and cost savings is impressive, as shown in the following examples:

The following graphic shows the steps that are necessary for creating and updating maps and data products using German-specific AAA© data:

Step 1: Automation consists of completely automated steps. These are visualization according to a signature catalog, generalization, automatic text placement and automatic incremental updating.
 
Step 2: Editing and Finishing is the editng step which include any remaining interactive cartographic editing where conflicts still exist, creating the map layouts, grids, automatic grid labelling and then configuring the map output and plots for quality assurance. Output can include .pdf, raster or data products.

One of the most critical steps in the automated process is that of incremental updating of map products. Updating can include data that is being fed into the system as a result of changes in the real world or as a result of manual editing that has taken place. Up until now, it has been a challenge to keep maps and data up to date while considering both new and manually edited data. With axpand, it is now possible to perform these tasks in one integrated process using automation.
 
Summary – Automated generalization allows for a significant reduction in map production and updating processes. axpand’s automation includes visualization of data according to a signature catalog, automatic generalization, automatic text placement and automatic recurring updating. The automation process is controlled through constraints. The results have proven to be consistent every time and reach a level of up to 80% correctness. This high quality automation reduces required manual intervention by cartographers to 20% of what it was before. As calculations from the field show,  the positive impact of axpand automation on time and cost savings are impressive.

Resources
LVermGeo Saxony-Anhalt, base data: NAS-Data according to ‚GeoInfoDok’-Version 6.0

Käuferle et al. (2010): Käuferle, Dominik: Automatisierte Landeskartenherstellung. [Automatic Production of National Maps] Geomatik Schweiz 5/2010, S.197. Automatisierte Landeskartenherstellung in der Schweiz. [Automatic Production of National Maps in Switzerland]  Kartographische Nachrichten [Cartography News]  5/2010, S.274. Automatisierte Landeskartenherstellung. [Automatic Production of National Maps] VDV Vermessung und Geoinformation Magazin [VDV Survey and Geoinformation Magazine] 5/2010, S.342.

Arbeitsgemeinschaft der Vermessungsverwaltungen der Länder der Bundesrepublik Deutschland (AdV) [Working Committee of the Surveying Authorities of the States of the Federal Republic of Germany (AdV)], ATKIS-Signaturenkatalog für die Digitale Topographische Karten [ATKIS Signature Catalog for Digital Topographic Maps]: http://www.adv-online.de/