Yet even with all of these new developments, the ceiling for our industry remains high. We’ve only just begun to scratch the surface of the relevant use cases. For instance, outside of the typical finance and government sectors, we’re beginning to see increased interest in geospatial intelligence from other industries, such as insurance.
One of the factors powering this increase in use cases is the advent of data sources that utilize more of the electromagnetic spectrum than traditional visual imagery. For instance, we’re seeing satellites launch now that image using synthetic aperture radar (SAR), and satellites that can detect radio waves. These types of data present new challenges in analysis, as they require different kinds of algorithms to detect change and patterns. However, it’s also been interesting to see how many of the techniques we’ve developed for analyzing traditional EO imagery, such as deep learning, are also applicable to these new data sources. This is an area where we’ll continue to see development in the coming year.
Another trend that will gain momentum in 2019 is the use of GPUs to analyze non-imagery datasets. Historically, CPUs have been used for most computing needs, while GPUs have been reserved for tasks that require greater processing power, such as imagery analysis. But today, even non-imagery geospatial datasets are becoming quite massive, which means that you’ll start to see GPUs being used more and more outside of traditional computer vision tasks.
Overall, 2019 is poised to be an exciting year for geospatial analytics, with new applications being developed and advances being made in existing areas. In fact, this is a field with so much innovation on the horizon that it’s impossible to predict for certain what new technologies will evolve in the next twelve months—but it will undoubtedly be exciting to watch!