The Geo-Wiki was established in 2010 in the Novel Data Ecosystems for Sustainability research group, located within the Advancing Systems Analysis Program at the International Institute for Applied Systems Analysis (IIASA) in Laxenburg, Austria. An early beta-version was developed in partnership with the University of Freiburg, Germany and the University of Wiener Neustadt, Austria.
The Geo-Wiki provides anyone with the means to engage in monitoring of the earth's surface by classifying satellite, drone or ground-level imagery. Data can be input via desktop or mobile devices, with campaigns and games used to incentivize input. These innovative techniques have been used to successfully integrate citizen-derived data sources with expert and authoritative data to address pressing policy-related questions (e.g. European environmental policy, SDG indicators and more).
Since its inception, Geo-Wiki has grown rapidly, with currently over 22,000 registered users having contributed more than 18 million image classifications from around the world. Furthermore, the Geo-Wiki toolbox has expanded to include numerous applications which help to address a variety of global challenges (e.g. land use change, food security, pollution and more). In addition, we have many ongoing research projects that rely on and further develop these applications, kindly supported by e.g. the European Commission, the European Space Agency and the Austrian Research Promotion Agency among others.
We apply AI techniques to a variety of our research challenges. In particular we have used deep learning recently to detect Amazon deforestation, building up a large image library via crowdsourcing to train various deep learning algorithms (https://www.sas.com/en_us/data-for-good/rainforest.html). In another application, we are testing machine learning algorithms for global modelling of Zenith Wet Delay based on GNSS measurements and meteorological data (https://www.camaliot.org/). Furthermore, we have created a free and open AI image library classification tool (https://pppui.cloud.geo-wiki.org/), a crowdsourcing platform for efficiently and intuitively classifying images for machine learning. In addition, we are part of the RapidAI4EO (https://rapidai4eo.eu/) Project which is advancing rapid and continuous land monitoring with state-of-the-art AI solutions.
We manage a suite of infrastructure to support our various desktop and mobile applications. The basic infrastructure relies on 16 cloud servers and five physical servers, located within the EU. A Kubernetes cluster sits on top of these for development purposes. We use the open-source PostgreSQL and POSTGIS database management software for hosting our spatial databases. Furthermore, for hosting Web Map Services we employ the open-source Geoserver and Mapserver platforms. A suite of Azure services are used to manage the entire system. We also rely on several Google cloud services (ie. Google Earth Engine, Google Street View). In addition, we utilize Unity software for most of our mobile applications, with Google Flutter supporting some additional mobile apps, and Xamarin for some legacy apps.
asdfdsafasdf