Our Director of Analytics David Breeding held an exclusive conversation with Google’s Kyle Campbell, Global Lead of Geo Imagery and Analysis (Google), and Jeremy Malczyk, Cloud Geographer. Together, they “unboxed” and explored everything you need to know about Google Earth Engine.
Highlights and Soundbites:
Kyle Campbell: “The real interesting thing is, we see a lot of companies in the cloud space starting to enter the remote sensing world. We have a very large ecosystem of users; so while Earth Engine was launched as a commercially available product back in June of last year, we’ve been around for over 12 years, and we have over 50,000 people monthly that come on the academia side, the research side, and now the business side to solve problems. This rich user community provides a ton of benefit as a commercial user if you have issues, questions – there’s a really great ecosystem of people to ask questions, and of course, Google is here to support you as well as our friends at Sanborn and AppGeo.”
Kyle: “What we like to say with Earth Engine [is] it helps you turn pixels into insights, and so, using all of this massive amount of data that we’ve collected for you. I would like to point out you don’t have to use just the data that’s there – our friends here at Sanborn collect a lot of aerial imagery. As long as the imagery you own maybe on-prem or on site or maybe in another cloud is geo-referenced, you can bring that to Earth Engine and have your own private instance. You can use that higher resolution data as well and and basically take that data, compute, analyze, and then create actionable reports and visualize it at the end. Really this is the workflow that we look at when we’re using Earth engine is taking the data and turning the pixels into insights for your business or your organization.”
David Breeding: “You can start to set up monitoring moments that say, ‘when there’s big deviations in what we were expecting, let me know about that,’ and you have (because of the depth of the Google Earth Engine Catalog) control over that that temporal unit that you’re interested in, and so that’s just real power to drive decision making, to validate assumptions, you can look back through the catalog and can really start to to get a grasp on some of the trends and and indices that are of interest to you.”
David: “R&D is a part of this as well. You need to make space for people to develop these things. You can pull things off the shelf, and there’s a lot of things that you can employ really quickly, but there’s also going to be as you go down this journey more opportunity to create new things. But, thinking about a basic implementation – there’s the data piece, there’s the analysis, and results piece, and then there’s that operational application piece, and this is how it starts.It doesn’t have to be that complex to start to create value.”
David: “We’re seeing some really interesting thoughts around how machine learning processes and the Earth Engine assets can be paired together to improve the amount of training information that’s easily accessible. Satellites continue to fly. There’s new imagery every day. We’d love to be able to profile that. We’d love to be able to predict on that, and so Earth Engine’s starting to play a role in feeding these machine learning engines. There’s Earth Engine assets, and then there’s also high resolution assets. Sanborn is capturing really high-res information we’ve been thinking about how that feathers in.”