More from Open Street Map…Thoughts on Data

The Open Street Map (OSM) community has accomplished great things during its relatively brief existence. Using state of the art mapping technology in an open source platform, OSM is tackling global issues head on, and the track record is very impressive.

OSM is currently being used in a large number of humanitarian projects such as malaria elimination, drought and famine relief, plus numerous natural disasters such as earthquakes, floods. For more details on humanitarian mapping projects, go to: https://www.hotosm.org/projects/disaster-mapping

OSM is enabling the development of freely-reusable geospatial data, which is an incredibly valuable resource. Geoffrey Kateregga, Lead Mapping Supervisor of OSM Africa, summed it up best when he said,

“Data is the oil of the 21st century.”

And OSM is immersed in the development of data. They have created several open source applications to leverage collaborative mapping. And they have expanded involvement in mapping by training volunteers to use these apps to contribute information to the map.

Because the management of data is so useful to so many urgent projects, OSM has a bright future. However, the community does have some data-related issues to resolve among themselves, and the issues are unfortunately inherent to OSM’s open structure and global scope.

Put most simply, there is just WAY too much data out there to record. The Earth is a big place with a lot of features, both natural and human. And all that data needs to be updated regularly, in some cases in real time. The questions OSM faces are, ‘How do we get all the data we need on OSM as fast, and accurately, as possible? And how do we set up a system to refresh that data consistently and reliably?’

Right now, the two proposals to answer those questions are: machine learning and volunteers.

Machine learning works like this. Say you want to map all the buildings in an area. First you ‘train’ an algorithm to recognize buildings by manually identifying buildings that represent what all the buildings in the area look like. The program analyzes imagery and identifies new buildings based on what looks like the initial inputs that you provided manually.

On the other hand, the idea behind volunteering is that the OSM community organizes mapping parties, called ‘mapathons’, to map an area together. Anyone can volunteer, with the more skilled users handling more advanced work and coaching the newbies on the finer points.

There are quality issues with both of these approaches. Volunteers are often inexperienced and can easily make errors. And machine learning has been known to identify features, like clouds or trees, as something they’re not.

These two approaches are by no means mutually exclusive, and both will likely continue to be used.

Many in the OSM community seem to be leaning towards machine learning. Computers provide greater reliability performing tedious head-pounding tasks, while volunteers often require a feeling of accomplishment (sometimes difficult to convey in data entry) and can be fickle. On the other hand, the volunteer system develops a cohort of people who are extremely interested in the OSM projects, and their enthusiasm serves to advance awareness of the organization.

MapAThon in Istanbul, Turkey

So whether the data is entered by hand or by machine, the OSM map can be used to solve a plethora of social, economic, and humanitarian issues, The data is just sitting there waiting to be leveraged.

So if you want to contribute to your local map, use geospatial data to solve problems in your community, or simply learn about another part of the world, check out www.openstreetmap.org.

Happy mapping!