GeoAI is already here but rapid advancements mean that its adoption is harder than it needs to be
It would be wrong to say that artificial intelligence and machine learning technologies have been quietly changing the geospatial information industry, since so many vendors have been keen to shout about their AI credentials at every opportunity, but the past decade has seen big changes brought about in how data is obtained and processed, thanks to innovations in this area.
So big has the transformation been, it has naturally acquired a name with a ‘geo’ prefix: GeoAI. However, as with all technologies that have expanded ‘under the radar’, different vendors have gone about their work in different ways. We’re now at a crossroads where the industry needs to pause, perhaps even stop slightly, so that vendors can work together on standards for integrating GeoAI technologies with each other, as well as different technologies, particularly so they can be used for different business processes without users having to start from scratch.
On page 30, Simon Chester and Kyoung-Sook Kim, a co-chair of the OGC GeoAI Domain Working Group, discuss what needs to be done to achieve this. In particular, machine intelligences are no different from their human counterparts – they learn from what they sense around them. Feed them bad data and their idea of the world will be skewed…
I hope you enjoy the issue.