Open Street Map(ping)

To mark the end of another experimental year, following is an account of first impressions of the OSM environment. A rudimentary understanding of digital cartography is helpful, however, armed with the descriptive notes, I believe the average reader will find it engaging as well. This in turn hopefully adds to the list of active OSM mappers, especially in India. Although a well-known and abundantly used platform around the world, India’s participation is somewhat limited. Its road data is extensive, but land-use demarcations and building footprints are sorely missing. Since it is openly sourced AND easy to use, most of us could consciously opt to contribute to the database. It does not take up a lot of time and third-party mobile applications for editing OSM, like Vespucci and StreetComplete, are readily available. On the go mapping may seem excessive for non-mappers but ensures correctness of data in lieu of very little effort. Although its uses range all the way from simple navigation to national humanitarian efforts, being from the construction industry myself, I find the repository extremely utilitarian when assessing neighbourhoods and settlement systems. To put it more accurately, I find the lack of data limiting during such studies. Government agencies are the largest singular source and there are visible efforts to digitize India’s socio-economic and geographic archives. With appropriate licenses in place, these could be shared on openly sourced platforms which, in my opinion, is the best route to expanding this database. The information may be exported from OSM in a variety of ways and could be used for anything from conceptual illustrations to building digital twins. The following article associates traditional mapping concepts and processes with the OSM ecosystem. It is also a reflection on the benefits of crowdsourcing GIS data and building a community around it. Our sample site here is the IIT (ISM) campus situated in Dhanbad, Jharkhand, India, primarily since it is well known to me. I trust the readership will find ample resources to navigate signing up, logging in, and reaching the OSM iD mapping interface (below).

Zooming in to an appropriate scale some of the main institute buildings can be seen marked. The geometry here does not neatly overlap with the satellite building footprint. Parallax may be the result of careless drafting or a different underlay. Choosing an appropriate image source with better positioning is a far more practical option than offsetting such a map. Existing features may be edited just as readily as new ones can be created. For the purposes of this article, I picked a comparatively unmarked area that happens to be predominantly residential. The default iD editor has been used for this run.

The three feature drawing options in the tools panel at the top mimic universal GIS. This is because of shapefiles (refer to the details panel if required). An appropriate type for each feature must therefore be selected.

Having marked a feature, one can see the ‘Edit Feature’ panel pop up to the left which collects metadata associated with each geometry. This is the attribute table (more in the details panel) of the OSM ecosystem, albeit far more flexible, and is its MOST essential component.

The first section is Feature Type which specifies the land-use, building, or natural feature type of the selected element. The next, Issues, is temporary but highlights an important aspect associated with OSM mapping. Every feature needs to have a ‘type’. The default, ‘Area’, is only a placeholder and is not considered acceptable. There are three other sections, namely Fields, Tags, and Relations. Fields comes in a pre-set combination depending on the Feature Type selected. So a ‘House’ feature type will not have the same set of fields as say a ‘Pond’ feature type (duh I guess). This helps ensure relevant information is collected. There is also the option to add more fields, but this is left at the discretion of the mapper. Tags are listed automatically based on the data provided but more may be added independently. Relations is used to associate the feature with surrounding features and geography. This is important for setting up the local and regional context of every geometry. It also displays parent-child relations when working with multi-polygons. Together these sections describe the feature in question which enables it to be used in a variety of ways. A small warning notification in the information panel at the bottom shows how many features drawn by the mapper have a missing type field. The geometry also looks different which is helpful. The Edit Feature panel is almost the same for point and line features as well, albeit with contextual differences (much like that between fields and feature types).

Outlining and editing a few other features, I then proceeded to ‘save’ the changes made (button on the top right corner). Adding a changeset comment is the most important, and mandatory, part of upload. It is essentially a description of the work being published. Although any text fulfils the requirement, it is expected to add value to changes made such that users may understand the data better. There is also the option to open your edits for review. This selection cannot be changed once published and opting for it, more often than not, prolongs the time taken for updates to appear online.

Information may be exported directly from OpenStreetMap or downloaded via third-party applications that aim to supplement the dataset. There is a multiplicity of such platforms, each serving a different purpose, too diverse to be included in this discussion. Direct downloads give files with .osm extension. This filetype is coded in XML and is exclusive to OpenStreetMap. It cannot be opened as is which is why most third-party applications convert it to a more compatible format. The reason for its disparate nature lies in its origin. The filetype was developed with the express intent of distributing cartographic data over the internet whereas other traditional, and professional, mapping and assessment tools are founded on their ability to support queries. OSM has been around for a long time now and consequently multiple applications and plugins exist that seamlessly convert it for further use.