Everywhere we turn these days, there’s yet another disruptive technology seemingly set to take the world by storm with increased automation and reduced cost. That’s certainly an apt description of unmanned aircraft systems (UASs) vs. manned aerial imaging; cloud servers vs. setting up and maintaining a high-performance computing center; and ever-more-capable smallsat Earth-observing constellations where tens and even hundreds of satellites cost less than the leading high-resolution commercial imaging satellite.
When such disruption takes place, it’s difficult to track all the players and their fortunes or combinations with existing players. Although this level of industry churn is disconcerting for those who like to know all that’s going on, the end result is a boon for practitioners who get more data sources, better data handling and analysis, and cheaper-yet-more-capable solutions.
One of the more-interesting offshoots of all this activity is the ability to image an area more easily than in the past. This certainly is true with UAS technology that can be readily deployed to map and monitor at far-more-frequent intervals. It’s also the case with satellite constellations that promise daily refreshes of any point on our planet.
Coupled with this is the ability to access such information on any device from most any location. Continuing improvements with mobile device data capacity are being matched by increasing wireless bandwidth that allows for higher-resolution imagery to be communicated quickly from device to device using commercially available bandwidth.
This combination of easy capture and quick sharing makes the Earth-observation toolset an interesting new input to a wide variety of business-related problem-solving and decision-support applications. This is perhaps the perfect convergence that will finally elevate Earth observation to meet its promise in the commercial market.
Patterns for Problems
The large increase in data volume is giving rise to insight-as-a-service providers who can analyze imagery quickly via machine-learning algorithms that decipher patterns to return insight—cloud computing provides the horsepower to achieve such analysis. The volume of data is the key ingredient, with insights only improving as patterns are refined, and pipelines of more and fresher imagery become available.
Querying large datasets quickly with high throughput may seem impossible for average practitioners, but these barriers are coming down. Microsoft recently launched Azure Machine Learning to make this level of data science accessible for those wishing to develop predictive analytical applications. IBM also offers Watson Developer Cloud with several cognitive services as RESTful APIs for similar outcomes.
Fresher data coming in at unprecedented volumes stands to disrupt many geospatial businesses while providing amazing new insights unimaginable a few years ago. Be certain to read this issue’s “Industry Outlook” to see what the future holds according to our Editorial Advisory Board, who are tackling these disruptions head-on and helping forge the future of Earth observation.