Recently I became an official member of the Humanitarian OpenStreetMap Team (HOT), a registered non-profit organization in the United States. Since over two years I'm an active contributer and participant in different OpenStreetMap (OSM) activities all over the world but with a special focus on Nicaragua, the country I'm living in. OpenStreetMap consists in an Open Data hub for geographic information and can be described as the “Wikipedia for geo information”.
The Humanitarian OpenStreetMap Team is based on the principle that “free map data would be a tremendous benefit for humanitarian aid and economic development” (OSM wiki about HOT, visited 26th of March 2015) and the main activity are the organization of activations of worldwide volunteers to collaborate all together over the OpenStreetMap platform to raise geographic data for disaster response after natural catastrophes (such as the recent Ebola outbreak 2014, the tsunami 2013 in the Philippines or the earthquake in Haiti 2010). This information is then freely available for everybody but in particular to organizations and government working in the field to safe people's life and improve the situation on the ground.
The OpenStreetMap project started in 2004 to build the first free world map. The area of interest of OpenStreetMap is the whole planet earth. The Humanitarian OpenStreetMap Team started in 2009, but became publicly visible and became attention in 2010 for the Haiti earthquake response.
The whole organization is based on the use of Geographical Information Systems and all parts of the organization are pushing the process because of the huge benefits for humanitarian and development issues. There can not be drawn a straight line between producers and users of the map information. Most of the actors are both consumers and creators of the hub of geographic information. Therefore there are no difficulties about acceptance of GIS at all, because this is common sense in the Humanitarian OpenStreetMap Team.
The basic methodology of raising data is done through a crowd-sourcing approach. Which means, that the crowd, or everybody with a certain level of computer knowledge can contribute to the map. This is done in the general way how OpenStreetMap works. On the website people can “edit” the geographic data through a simple editor (Illustration 2). There are other editors, such as the desktop application JOSM, and tools, such as an API interface to apply python scripts, available to edit OpenStreetMap's data in a more advanced way.
Crowd-sourcing means, that all levels of expertise are welcome and taken into consideration. The same principle also applies for the knowledge, where local knowledge is mostly the requirement, but also the guarantee for good quality of the data. This unconventional data collecting method by using a crowd-sourcing approach has led to an overall database of geographic information of the world, counting of an amount of over 550GB of data (Wiki about planet download, visited 26.03.2015). All this information is publicly available to anybody and for any purpose, licensed under the Commons Open Database License” (ODbL).
This allows the following:
- Free sharing (copying, distribution and use) of the database.
- Permits the creation and productions of works from the database.
- Allows the adaption (modification, transformation and building upon) of the database.
Under the following conditions:
- Attribution (the OpenStreetMap project has to be attributed when using the data).
- Share-Alike: If publicly used the result must be under the same or compatible license.
- Prohibits exclusive use in restricted systems (such as DRM).
These principle is also known as Open Data and has been used extensively and successfully in the field of Open Source and Free Software since the nineteen-seventies and it is comparable to the license of the Wikipedia project. All data can be downloaded, either in a complete way for the whole planet, or as country extracts, or with the use of certain tools, like Overpass API, the database of geographic information can be queried to obtain, extract and download only desired parts of the data available.
The Humanitarian OpenStreetMap Team uses OpenStreetMap to provide geographic information in places of the world when they are most needed. Generally this happens through so called “activations”, which aim to invite volunteers from all parts of the world. These are usually publicity campaigns over the medias to involve people that want to help, to map remotely by the use of satellite imagery a certain part of the world, where a natural catastrophe, such as a flooding, and earthquake, but also social crisis has happened and a need for map data is present.
In order to coordinate the work, an Open Source web application called the OSM Tasking Manager (Illustration 3) has been developed by the organization. The specific area of interest is added as a “project” and a description of the task, what to map, is given. Then this area gets divided into small squares and volunteers can block one square to advice other people, that he or she is mapping in this part and to avoid conflicts with others when saving the changes. The person completes the task on this part, the selected square and gives feedback in the OSM Tasking Manager if the editing was completed, or if the person encountered problems and somebody else should have a look on this part.
The data raised remotely through the use of satellite imagery is reduced to the observable objects in such images. This means, that streets and water bodies are the first tasks for remote sensing. This data needs then to be completed by people with local knowledge or individuals that are on the ground and can verify names and type of objects.
The outcome of such activations that organizes the Humanitarian OpenStreetMap Team is impressive as it can be examined for the Haiti earthquake response activation (see Illustration 4). The power of collaborative work can be observed in each one of twenty-nine realized the GIS activities. At this moment there are six ongoing activations.
The amount of mostly reliable information obtained through this crowd-sourcing method is huge and very quick in comparison with traditional data surveying, note the level of detail in the map of Illustration 5.
Of course, the nature of the attribute data is not very hierarchically structured nor always very consistent, because many people, from different cultures and parts of the world and therefore many ways of classifying objects are moving this gigantic amount of geographic information. Nevertheless the central documentation point is the Wiki application of OpenStreetMap, where classification of objects underlay a constant forming process. Automatic scripts help to achieve as much as possible constancy in the world-wide data. And the standard rendering on the website helps people to verify their valid classification. In general the principle of many eyes building up the information together leads to a satisfying result.
The data is then available under the legal conditions described earlier. But also the data is available in open and documented formats, which allow a wide use of the data. See the illustration 5, of the data used by rescue teams in Haiti on their GPS device.
A wide range of different organizations and governmental institutions have been relying on the data collected through the Humanitarian OpenStreetMap Team activations.
This data is used on GPS devices, smart phone applications and also on printed paper maps, which can be easily obtained with available Open Source tools. The data can also be downloaded in vector formats, which allow the integration into other Geographic Information Systems and the data can be used to make analysis or design maps with a certain purpose.
The amount of results have been overwhelming, for the twenty-nine activations happened until now. All freely available for anybody in the world biggest geographic information hub. Review the list of the ongoing and past activations of the organization all over the world.
The unconventional method of data collection implies a lot of opportunities, but also big difficulties and challenges. Probably the greatest difficulties are the consistency of data classification and the completeness of information. It is common to have certain parts with very good geographic information and other, seemingly less interesting, but not necessarily less important, parts where the information becomes incomplete or even false in very few occasions.
The Humanitarian OpenStreetMap Team does not take into consideration of the decision process, the origin or field of local knowledge. This is done through communication, but could potentially be taken into consideration use GIS methods.
The aim of the organization is not only to provide free map data for humanitarian aid purposes, but also for opportunities of economic development. So, there is still a good work to do, to find a broader acceptance and usage of the information for more than the most urgent needs, as disaster response.
More development is being applied to the line of automatic treatment of the data, especially automatic error detection (for example: two lines that cross, but do not connect or are not categorized as a bridge). These errors are more difficult to find manually, but can then be fixed easily by having a human eye looking at it.
For other errors, such as bad names in the attributes, or changes made to the classification can be even applied automatically. This is still a field that has to be explored more in the Humanitarian OpenStreetMap Team and the OpenStreetMap project in general.
Another great difficulty is the interoperability of data. OpenStreetMap, as a bottom-up project builds up classification of data in a collaborative way, which is very complex taking into consideration all views of people from different disciplines and parts of the world. A good challenge is to develop a recognized standard on geographic object classification, which would then permit the data to be more exchangeable in the future.
The Humanitarian OpenStreetMap Team uses GIS methods extensively to make the world a better place and to safe lives. The main challenges are more in the social part of getting people involved into the data collection, definition and maintaining process, than in the pure GIS techniques.
This report has been elaborated for the course “GISA21 – Introduction to Geographical Information Systems” within the Master's program iGEON at Lund University, Sweden in V2015 (Spring).